diff --git a/fern/.gitignore b/fern/.gitignore
new file mode 100644
index 0000000000..2114f1dd17
--- /dev/null
+++ b/fern/.gitignore
@@ -0,0 +1,2 @@
+**/.preview
+**/.definition
diff --git a/fern/fern.config.json b/fern/fern.config.json
new file mode 100644
index 0000000000..e6d3226d4c
--- /dev/null
+++ b/fern/fern.config.json
@@ -0,0 +1,4 @@
+{
+ "organization": "nvidia",
+ "version": "5.46.1"
+}
\ No newline at end of file
diff --git a/fern/generators.yml b/fern/generators.yml
new file mode 100644
index 0000000000..746dd83b81
--- /dev/null
+++ b/fern/generators.yml
@@ -0,0 +1,20 @@
+# yaml-language-server: $schema=https://schema.buildwithfern.dev/generators-yml.json
+api:
+ specs:
+ - openapi: ../openapi/openapi.yaml
+default-group: local
+groups:
+ local:
+ generators:
+ - name: fern-typescript-sdk
+ output:
+ location: local-file-system
+ path: ../sdks/typescript
+ version: 3.71.8
+ python-sdk:
+ generators:
+ - name: fern-python-sdk
+ version: 5.14.14
+ output:
+ location: local-file-system
+ path: ../sdks/python
diff --git a/sdks/python/.fern/metadata.json b/sdks/python/.fern/metadata.json
new file mode 100644
index 0000000000..79a21230c8
--- /dev/null
+++ b/sdks/python/.fern/metadata.json
@@ -0,0 +1,8 @@
+{
+ "cliVersion": "5.46.1",
+ "generatorName": "fernapi/fern-python-sdk",
+ "generatorVersion": "5.14.14",
+ "originGitCommit": "36bbf2906e7b4ada998c6006932f83e2399e9ac2",
+ "originGitCommitIsDirty": true,
+ "invokedBy": "manual"
+}
\ No newline at end of file
diff --git a/sdks/python/CONTRIBUTING.md b/sdks/python/CONTRIBUTING.md
new file mode 100644
index 0000000000..af948ce725
--- /dev/null
+++ b/sdks/python/CONTRIBUTING.md
@@ -0,0 +1,125 @@
+# Contributing
+
+Thanks for your interest in contributing to this SDK! This document provides guidelines for contributing to the project.
+
+## Getting Started
+
+### Prerequisites
+
+- Python 3.9+
+- pip
+- poetry
+
+### Installation
+
+Install the project dependencies:
+
+```bash
+poetry install
+```
+
+### Building
+
+Build the project:
+
+```bash
+poetry build
+```
+
+### Testing
+
+Run the test suite:
+
+```bash
+poetry run pytest
+```
+
+### Linting and Formatting
+
+Check code style:
+
+```bash
+poetry run ruff check .
+poetry run ruff format .
+```
+
+### Type Checking
+
+Run the type checker:
+
+```bash
+poetry run mypy .
+```
+
+## About Generated Code
+
+**Important**: Most files in this SDK are automatically generated by [Fern](https://buildwithfern.com) from the API definition. Direct modifications to generated files will be overwritten the next time the SDK is generated.
+
+### Generated Files
+
+The following directories contain generated code:
+- `src/` - API client classes and types
+- Most Python files in the project
+
+### How to Customize
+
+If you need to customize the SDK, you have two options:
+
+#### Option 1: Use `.fernignore`
+
+For custom code that should persist across SDK regenerations:
+
+1. Create a `.fernignore` file in the project root
+2. Add file patterns for files you want to preserve (similar to `.gitignore` syntax)
+3. Add your custom code to those files
+
+Files listed in `.fernignore` will not be overwritten when the SDK is regenerated.
+
+For more information, see the [Fern documentation on custom code](https://buildwithfern.com/learn/sdks/overview/custom-code).
+
+#### Option 2: Contribute to the Generator
+
+If you want to change how code is generated for all users of this SDK:
+
+1. The Python SDK generator lives in the [Fern repository](https://github.com/fern-api/fern)
+2. Generator code is located at `generators/python-v2/`
+3. Follow the [Fern contributing guidelines](https://github.com/fern-api/fern/blob/main/CONTRIBUTING.md)
+4. Submit a pull request with your changes to the generator
+
+This approach is best for:
+- Bug fixes in generated code
+- New features that would benefit all users
+- Improvements to code generation patterns
+
+## Making Changes
+
+### Workflow
+
+1. Create a new branch for your changes
+2. Make your modifications
+3. Run tests to ensure nothing breaks: `poetry run pytest`
+4. Run linting and formatting: `poetry run ruff check .` and `poetry run ruff format .`
+5. Run type checking: `poetry run mypy .`
+6. Build the project: `poetry build`
+7. Commit your changes with a clear commit message
+8. Push your branch and create a pull request
+
+### Commit Messages
+
+Write clear, descriptive commit messages that explain what changed and why.
+
+### Code Style
+
+This project uses automated code formatting and linting. Run `poetry run ruff format .` and `poetry run ruff check .` before committing to ensure your code meets the project's style guidelines.
+
+## Questions or Issues?
+
+If you have questions or run into issues:
+
+1. Check the [Fern documentation](https://buildwithfern.com)
+2. Search existing [GitHub issues](https://github.com/fern-api/fern/issues)
+3. Open a new issue if your question hasn't been addressed
+
+## License
+
+By contributing to this project, you agree that your contributions will be licensed under the same license as the project.
diff --git a/sdks/python/__init__.py b/sdks/python/__init__.py
new file mode 100644
index 0000000000..2130908017
--- /dev/null
+++ b/sdks/python/__init__.py
@@ -0,0 +1,1397 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
+import typing
+from importlib import import_module
+
+if typing.TYPE_CHECKING:
+ from .types import (
+ ActionRails,
+ ActivatedRail,
+ Adapter,
+ AdapterEntityFilter,
+ AdapterEntityFilterDescription,
+ AdapterEntityFilterModel,
+ AdapterEntityFilterName,
+ AdaptersPage,
+ AiDefenseRailConfig,
+ Annotation,
+ AnnotationFilter,
+ AnnotationInput,
+ AnnotationInput_Feedback,
+ AnnotationInput_Label,
+ AnnotationInput_Metadata,
+ AnnotationInput_Note,
+ AnnotationKind,
+ AnnotationSortField,
+ Annotation_Feedback,
+ Annotation_Label,
+ Annotation_Metadata,
+ Annotation_Note,
+ AnnotationsPage,
+ ApiEndpointData,
+ AtifAgent,
+ AtifContentPart,
+ AtifContentPartImage,
+ AtifContentPartText,
+ AtifContentPart_Image,
+ AtifContentPart_Text,
+ AtifFinalMetrics,
+ AtifImageSource,
+ AtifImageSourceMediaType,
+ AtifMetrics,
+ AtifObservation,
+ AtifObservationResult,
+ AtifObservationResultContent,
+ AtifStep,
+ AtifStepAgent,
+ AtifStepAgentMessage,
+ AtifStepAgentReasoningEffort,
+ AtifStepSystem,
+ AtifStepSystemMessage,
+ AtifStepUser,
+ AtifStepUserMessage,
+ AtifStep_Agent,
+ AtifStep_System,
+ AtifStep_User,
+ AtifSubagentTrajectoryRef,
+ AtifToolCall,
+ AuthContext,
+ AuthDiscoveryResponse,
+ AutoAlignOptions,
+ AutoAlignRailConfig,
+ BackendFormat,
+ BaseModelFilter,
+ BaseModelFilterName,
+ CacheStatsConfig,
+ CacheStatus,
+ CapturedChatCompletionsRequest,
+ CapturedChatCompletionsResponse,
+ CapturedChatMessage,
+ ChatCompletionAssistantMessageParam,
+ ChatCompletionContentPartImageParam,
+ ChatCompletionContentPartTextParam,
+ ChatCompletionFunctionMessageParam,
+ ChatCompletionMessageToolCallParam,
+ ChatCompletionMessageToolCallParamType,
+ ChatCompletionSystemMessageParam,
+ ChatCompletionToolMessageParam,
+ ChatCompletionUserMessageParam,
+ ChatCompletionUserMessageParamContent,
+ ChatCompletionUserMessageParamContentOneItem,
+ ChatCompletionUserMessageParamContentOneItem_ImageUrl,
+ ChatCompletionUserMessageParamContentOneItem_Text,
+ ChatCompletionsIngestResponse,
+ ChatMessageRole,
+ ClavataRailConfig,
+ ClavataRailConfigLabelMatchLogic,
+ ClavataRailOptions,
+ ComputeResourceSpec,
+ ComputeResources,
+ ContainerExecutorConfig,
+ ContainerSpec,
+ ContentSafetyConfig,
+ CpuExecutionProviderInput,
+ CpuExecutionProviderOutput,
+ CrowdStrikeAidrRailConfig,
+ DatasetMetadataContent,
+ DatasetMetadataContentSchema,
+ DatasetMetadataContentSchemasByPathValue,
+ DateRangeFilter,
+ DatetimeFilter,
+ DeleteResponse,
+ DialogRails,
+ DistributedGpuExecutionProviderInput,
+ DistributedGpuExecutionProviderOutput,
+ DockerJobExecutionProfile,
+ DockerJobExecutionProfileConfig,
+ DockerJobNetworkConfig,
+ DockerJobStorageConfig,
+ DockerVolumeMount,
+ DockerVolumeMountKind,
+ E2EJobExecutionProfile,
+ Engine,
+ EntitiesPage,
+ Entity,
+ EvaluationContext,
+ EvaluatorAggregate,
+ EvaluatorResult,
+ EvaluatorResultDataType,
+ EvaluatorResultFilter,
+ EvaluatorResultSortField,
+ EvaluatorResultsPage,
+ ExecutedAction,
+ ExperimentContext,
+ ExperimentFilter,
+ ExperimentGroupFilter,
+ ExperimentGroupRequest,
+ ExperimentGroupResponse,
+ ExperimentGroupResponsesPage,
+ ExperimentRequest,
+ ExperimentResponse,
+ ExperimentResponsesPage,
+ ExperimentSessionFilter,
+ ExperimentSessionResponse,
+ ExperimentSessionResponsesPage,
+ FactCheckingRailConfig,
+ FeedbackAnnotation,
+ FeedbackAnnotationInput,
+ FeedbackAnnotationInputValue,
+ FeedbackAnnotationValue,
+ FiddlerGuardrails,
+ FileStorageType,
+ FilesetFileOutput,
+ FilesetFilter,
+ FilesetFilterDescription,
+ FilesetFilterName,
+ FilesetMetadataInput,
+ FilesetMetadataOutput,
+ FilesetOutput,
+ FilesetOutputStorage,
+ FilesetOutputStorage_Huggingface,
+ FilesetOutputStorage_Local,
+ FilesetOutputStorage_Ngc,
+ FilesetOutputStorage_S3,
+ FilesetOutputsPage,
+ FilesetPurpose,
+ FinetuningType,
+ FinetuningTypeFilter,
+ FloatFilter,
+ Function,
+ FunctionCall,
+ GLiNerDetection,
+ GLiNerDetectionOptions,
+ GenerationLog,
+ GenerationLogOptions,
+ GenerationOptions,
+ GenerationOptionsOutputVars,
+ GenerationRailsOptions,
+ GenerationRailsOptionsInput,
+ GenerationRailsOptionsOutput,
+ GenerationRailsOptionsRetrieval,
+ GenerationStats,
+ GenericSortField,
+ GpuExecutionProviderInput,
+ GpuExecutionProviderOutput,
+ GuardrailCheckResponse,
+ GuardrailConfig,
+ GuardrailConfigData,
+ GuardrailConfigFilter,
+ GuardrailConfigFilterDescription,
+ GuardrailConfigFilterName,
+ GuardrailConfigsPage,
+ GuardrailsAiRailConfig,
+ GuardrailsAiValidatorConfig,
+ GuardrailsDataInput,
+ GuardrailsDataInputConfig,
+ GuardrailsDataOutput,
+ HttpValidationError,
+ HuggingfaceStorageConfig,
+ HuggingfaceStorageConfigRepoType,
+ ImagePullSecret,
+ ImageUrl,
+ ImageUrlDetail,
+ InferenceParams,
+ IngestResponse,
+ InjectionDetection,
+ InputRails,
+ Instruction,
+ JailbreakDetectionConfig,
+ JobExecutionProfileConfig,
+ K8SNimOperatorConfig,
+ KubernetesEmptyDirVolume,
+ KubernetesJobExecutionProfile,
+ KubernetesJobExecutionProfileConfig,
+ KubernetesJobStorageConfig,
+ KubernetesObjectMetadata,
+ KubernetesPersistentVolumeClaim,
+ KubernetesVolume,
+ KubernetesVolumeMount,
+ LabelAnnotation,
+ LabelAnnotationInput,
+ LabelAnnotationInputValue,
+ LabelAnnotationInputValueType,
+ LabelAnnotationValue,
+ LabelAnnotationValueType,
+ LinearLayerSpec,
+ ListFilesetFilesResponse,
+ LlmCallInfo,
+ LocalStorageConfig,
+ LogAdapterConfig,
+ Lora,
+ MambaConfig,
+ MessageTemplate,
+ MetadataAnnotation,
+ MetadataAnnotationInput,
+ MiddlewareCall,
+ MoEConfig,
+ Model,
+ ModelCacheConfig,
+ ModelDeployment,
+ ModelDeploymentConfig,
+ ModelDeploymentConfigFilter,
+ ModelDeploymentConfigFilterDescription,
+ ModelDeploymentConfigFilterName,
+ ModelDeploymentConfigModelSpec,
+ ModelDeploymentConfigsPage,
+ ModelDeploymentFilter,
+ ModelDeploymentFilterConfig,
+ ModelDeploymentFilterName,
+ ModelDeploymentFilterStatusMessage,
+ ModelDeploymentStatus,
+ ModelDeploymentStatusHistoryItem,
+ ModelDeploymentsPage,
+ ModelEntity,
+ ModelEntityFilter,
+ ModelEntityFilterAdapters,
+ ModelEntityFilterBaseModel,
+ ModelEntityFilterDescription,
+ ModelEntityFilterFinetuningType,
+ ModelEntityFilterName,
+ ModelEntitySortField,
+ ModelEntitysPage,
+ ModelMetadataContent,
+ ModelMode,
+ ModelParameters,
+ ModelProvider,
+ ModelProviderFilter,
+ ModelProviderFilterDescription,
+ ModelProviderFilterHostUrl,
+ ModelProviderFilterName,
+ ModelProviderSort,
+ ModelProviderStatus,
+ ModelProvidersPage,
+ ModelSpec,
+ ModelType,
+ MultilingualConfig,
+ NgcStorageConfig,
+ NgcStorageConfigTargetType,
+ NoteAnnotation,
+ NoteAnnotationInput,
+ NumericFilter,
+ OidcDiscoveryResponse,
+ OpenAiListModelsResp,
+ OpenAiModelResp,
+ OtelExportLogsPartialSuccess,
+ OtelExportLogsServiceResponse,
+ OutputRails,
+ OutputRailsStreamingConfig,
+ PaginationData,
+ PangeaRailConfig,
+ PangeaRailOptions,
+ PatronusEvaluateApiParams,
+ PatronusEvaluateConfigInput,
+ PatronusEvaluateConfigOutput,
+ PatronusEvaluationSuccessStrategy,
+ PatronusRailConfigInput,
+ PatronusRailConfigOutput,
+ PlatformJobEnvironmentVariable,
+ PlatformJobListResultResponse,
+ PlatformJobListTaskResponse,
+ PlatformJobLog,
+ PlatformJobLogPage,
+ PlatformJobResponse,
+ PlatformJobResponsesPage,
+ PlatformJobResultResponse,
+ PlatformJobSecretEnvironmentVariableRef,
+ PlatformJobSortField,
+ PlatformJobSpecInput,
+ PlatformJobSpecOutput,
+ PlatformJobStatus,
+ PlatformJobStatusResponse,
+ PlatformJobStep,
+ PlatformJobStepSpecInput,
+ PlatformJobStepSpecInputExecutor,
+ PlatformJobStepSpecInputExecutor_Cpu,
+ PlatformJobStepSpecInputExecutor_Gpu,
+ PlatformJobStepSpecInputExecutor_GpuDistributed,
+ PlatformJobStepSpecInputExecutor_Subprocess,
+ PlatformJobStepSpecOutput,
+ PlatformJobStepSpecOutputExecutor,
+ PlatformJobStepSpecOutputExecutor_Cpu,
+ PlatformJobStepSpecOutputExecutor_Gpu,
+ PlatformJobStepSpecOutputExecutor_GpuDistributed,
+ PlatformJobStepSpecOutputExecutor_Subprocess,
+ PlatformJobStepStatusResponse,
+ PlatformJobStepWithContext,
+ PlatformJobStepWithContextsPage,
+ PlatformJobStepsListFilter,
+ PlatformJobTask,
+ PlatformJobTaskStatusResponse,
+ PlatformJobsListFilter,
+ PlatformJobsListFilterName,
+ PlatformJobsListFilterSource,
+ PlatformJobsListFilterStatus,
+ PlatformSecretAccessResponse,
+ PlatformSecretAdminRotationResponse,
+ PlatformSecretResponse,
+ PlatformSecretResponsesPage,
+ PrivateAiDetection,
+ PrivateAiDetectionOptions,
+ Project,
+ ProjectSortField,
+ ProjectsPage,
+ PromptData,
+ RailStatus,
+ RailsConfigDataInput,
+ RailsConfigDataOutput,
+ RailsConfigInput,
+ RailsConfigOutput,
+ RailsInput,
+ RailsOutput,
+ ReasoningConfig,
+ RegexDetection,
+ RegexDetectionOptions,
+ RetrievalRails,
+ RoleBinding,
+ RoleBindingFilter,
+ RoleBindingFilterGrantedBy,
+ RoleBindingFilterPrincipal,
+ RoleBindingFilterRole,
+ RoleBindingsPage,
+ S3StorageConfig,
+ S3StorageConfigSignatureVersion,
+ SecretRef,
+ SensitiveDataDetection,
+ SensitiveDataDetectionOptions,
+ ServedModelMapping,
+ SingleCallConfig,
+ SlidingWindowConfig,
+ Span,
+ SpanEvaluationContext,
+ SpanFilter,
+ SpanKind,
+ SpanSortField,
+ SpanStatus,
+ SpansPage,
+ StatusEnum,
+ StepLifecycle,
+ StorageConfigType,
+ StringFilter,
+ SubprocessExecutionProvider,
+ SubprocessJobExecutionProfile,
+ SubprocessJobExecutionProfileConfig,
+ SubprocessJobExecutionProfileProvider,
+ TaskPrompt,
+ TaskPromptMessagesItem,
+ ToolCallConfig,
+ ToolCallingMetadataContent,
+ ToolInputRails,
+ ToolOutputRails,
+ Trace,
+ TraceFilter,
+ TraceSortField,
+ TracesPage,
+ TracingConfig,
+ TrendMicroRailConfig,
+ UpdateAdapterRequest,
+ UserMessagesConfig,
+ ValidationError,
+ ValidationErrorLocItem,
+ VirtualModel,
+ VirtualModelInferenceConfig,
+ VirtualModelsPage,
+ VolcanoJobExecutionProfile,
+ VolcanoJobExecutionProfileConfig,
+ Workspace,
+ WorkspaceMember,
+ WorkspaceMemberListResponse,
+ WorkspacesPage,
+ )
+ from .errors import (
+ BadRequestError,
+ ConflictError,
+ InternalServerError,
+ NotFoundError,
+ ServiceUnavailableError,
+ UnprocessableEntityError,
+ )
+ from . import (
+ adapters,
+ annotations,
+ discovery,
+ entity_store,
+ evaluator_results,
+ experiment_groups,
+ experiments,
+ files,
+ guardrails,
+ iam,
+ inference_gateway,
+ ingest,
+ jobs,
+ model_deployment_configs,
+ model_deployments,
+ model_providers,
+ models,
+ otlp,
+ secrets,
+ secrets_admin,
+ spans,
+ traces,
+ virtual_models,
+ )
+ from ._default_clients import DefaultAioHttpClient, DefaultAsyncHttpxClient
+ from .client import AsyncNvidiaApi, NvidiaApi
+ from .experiment_groups import ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort
+ from .experiments import ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort
+ from .files import (
+ CreateFilesetRequestStorage,
+ CreateFilesetRequestStorage_Huggingface,
+ CreateFilesetRequestStorage_Local,
+ CreateFilesetRequestStorage_Ngc,
+ CreateFilesetRequestStorage_S3,
+ )
+ from .guardrails import (
+ GuardrailCheckRequestFunctionCall,
+ GuardrailCheckRequestMessagesItem,
+ GuardrailCheckRequestMessagesItem_Assistant,
+ GuardrailCheckRequestMessagesItem_Function,
+ GuardrailCheckRequestMessagesItem_System,
+ GuardrailCheckRequestMessagesItem_Tool,
+ GuardrailCheckRequestMessagesItem_User,
+ GuardrailCheckRequestStop,
+ GuardrailCheckRequestToolChoice,
+ )
+ from .ingest import AtifIngestRequestSchemaVersion
+ from .jobs import (
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem,
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Docker,
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_E2E,
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_KubernetesJob,
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Subprocess,
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_VolcanoJob,
+ )
+ from .spans import ListSpansApisIntakeV2WorkspacesWorkspaceSpansGetRequestMode
+ from .traces import (
+ GetTraceApisIntakeV2WorkspacesWorkspaceTracesIdGetRequestMode,
+ ListTracesApisIntakeV2WorkspacesWorkspaceTracesGetRequestMode,
+ )
+_dynamic_imports: typing.Dict[str, str] = {
+ "ActionRails": ".types",
+ "ActivatedRail": ".types",
+ "Adapter": ".types",
+ "AdapterEntityFilter": ".types",
+ "AdapterEntityFilterDescription": ".types",
+ "AdapterEntityFilterModel": ".types",
+ "AdapterEntityFilterName": ".types",
+ "AdaptersPage": ".types",
+ "AiDefenseRailConfig": ".types",
+ "Annotation": ".types",
+ "AnnotationFilter": ".types",
+ "AnnotationInput": ".types",
+ "AnnotationInput_Feedback": ".types",
+ "AnnotationInput_Label": ".types",
+ "AnnotationInput_Metadata": ".types",
+ "AnnotationInput_Note": ".types",
+ "AnnotationKind": ".types",
+ "AnnotationSortField": ".types",
+ "Annotation_Feedback": ".types",
+ "Annotation_Label": ".types",
+ "Annotation_Metadata": ".types",
+ "Annotation_Note": ".types",
+ "AnnotationsPage": ".types",
+ "ApiEndpointData": ".types",
+ "AsyncNvidiaApi": ".client",
+ "AtifAgent": ".types",
+ "AtifContentPart": ".types",
+ "AtifContentPartImage": ".types",
+ "AtifContentPartText": ".types",
+ "AtifContentPart_Image": ".types",
+ "AtifContentPart_Text": ".types",
+ "AtifFinalMetrics": ".types",
+ "AtifImageSource": ".types",
+ "AtifImageSourceMediaType": ".types",
+ "AtifIngestRequestSchemaVersion": ".ingest",
+ "AtifMetrics": ".types",
+ "AtifObservation": ".types",
+ "AtifObservationResult": ".types",
+ "AtifObservationResultContent": ".types",
+ "AtifStep": ".types",
+ "AtifStepAgent": ".types",
+ "AtifStepAgentMessage": ".types",
+ "AtifStepAgentReasoningEffort": ".types",
+ "AtifStepSystem": ".types",
+ "AtifStepSystemMessage": ".types",
+ "AtifStepUser": ".types",
+ "AtifStepUserMessage": ".types",
+ "AtifStep_Agent": ".types",
+ "AtifStep_System": ".types",
+ "AtifStep_User": ".types",
+ "AtifSubagentTrajectoryRef": ".types",
+ "AtifToolCall": ".types",
+ "AuthContext": ".types",
+ "AuthDiscoveryResponse": ".types",
+ "AutoAlignOptions": ".types",
+ "AutoAlignRailConfig": ".types",
+ "BackendFormat": ".types",
+ "BadRequestError": ".errors",
+ "BaseModelFilter": ".types",
+ "BaseModelFilterName": ".types",
+ "CacheStatsConfig": ".types",
+ "CacheStatus": ".types",
+ "CapturedChatCompletionsRequest": ".types",
+ "CapturedChatCompletionsResponse": ".types",
+ "CapturedChatMessage": ".types",
+ "ChatCompletionAssistantMessageParam": ".types",
+ "ChatCompletionContentPartImageParam": ".types",
+ "ChatCompletionContentPartTextParam": ".types",
+ "ChatCompletionFunctionMessageParam": ".types",
+ "ChatCompletionMessageToolCallParam": ".types",
+ "ChatCompletionMessageToolCallParamType": ".types",
+ "ChatCompletionSystemMessageParam": ".types",
+ "ChatCompletionToolMessageParam": ".types",
+ "ChatCompletionUserMessageParam": ".types",
+ "ChatCompletionUserMessageParamContent": ".types",
+ "ChatCompletionUserMessageParamContentOneItem": ".types",
+ "ChatCompletionUserMessageParamContentOneItem_ImageUrl": ".types",
+ "ChatCompletionUserMessageParamContentOneItem_Text": ".types",
+ "ChatCompletionsIngestResponse": ".types",
+ "ChatMessageRole": ".types",
+ "ClavataRailConfig": ".types",
+ "ClavataRailConfigLabelMatchLogic": ".types",
+ "ClavataRailOptions": ".types",
+ "ComputeResourceSpec": ".types",
+ "ComputeResources": ".types",
+ "ConflictError": ".errors",
+ "ContainerExecutorConfig": ".types",
+ "ContainerSpec": ".types",
+ "ContentSafetyConfig": ".types",
+ "CpuExecutionProviderInput": ".types",
+ "CpuExecutionProviderOutput": ".types",
+ "CreateFilesetRequestStorage": ".files",
+ "CreateFilesetRequestStorage_Huggingface": ".files",
+ "CreateFilesetRequestStorage_Local": ".files",
+ "CreateFilesetRequestStorage_Ngc": ".files",
+ "CreateFilesetRequestStorage_S3": ".files",
+ "CrowdStrikeAidrRailConfig": ".types",
+ "DatasetMetadataContent": ".types",
+ "DatasetMetadataContentSchema": ".types",
+ "DatasetMetadataContentSchemasByPathValue": ".types",
+ "DateRangeFilter": ".types",
+ "DatetimeFilter": ".types",
+ "DefaultAioHttpClient": "._default_clients",
+ "DefaultAsyncHttpxClient": "._default_clients",
+ "DeleteResponse": ".types",
+ "DialogRails": ".types",
+ "DistributedGpuExecutionProviderInput": ".types",
+ "DistributedGpuExecutionProviderOutput": ".types",
+ "DockerJobExecutionProfile": ".types",
+ "DockerJobExecutionProfileConfig": ".types",
+ "DockerJobNetworkConfig": ".types",
+ "DockerJobStorageConfig": ".types",
+ "DockerVolumeMount": ".types",
+ "DockerVolumeMountKind": ".types",
+ "E2EJobExecutionProfile": ".types",
+ "Engine": ".types",
+ "EntitiesPage": ".types",
+ "Entity": ".types",
+ "EvaluationContext": ".types",
+ "EvaluatorAggregate": ".types",
+ "EvaluatorResult": ".types",
+ "EvaluatorResultDataType": ".types",
+ "EvaluatorResultFilter": ".types",
+ "EvaluatorResultSortField": ".types",
+ "EvaluatorResultsPage": ".types",
+ "ExecutedAction": ".types",
+ "ExperimentContext": ".types",
+ "ExperimentFilter": ".types",
+ "ExperimentGroupFilter": ".types",
+ "ExperimentGroupRequest": ".types",
+ "ExperimentGroupResponse": ".types",
+ "ExperimentGroupResponsesPage": ".types",
+ "ExperimentRequest": ".types",
+ "ExperimentResponse": ".types",
+ "ExperimentResponsesPage": ".types",
+ "ExperimentSessionFilter": ".types",
+ "ExperimentSessionResponse": ".types",
+ "ExperimentSessionResponsesPage": ".types",
+ "FactCheckingRailConfig": ".types",
+ "FeedbackAnnotation": ".types",
+ "FeedbackAnnotationInput": ".types",
+ "FeedbackAnnotationInputValue": ".types",
+ "FeedbackAnnotationValue": ".types",
+ "FiddlerGuardrails": ".types",
+ "FileStorageType": ".types",
+ "FilesetFileOutput": ".types",
+ "FilesetFilter": ".types",
+ "FilesetFilterDescription": ".types",
+ "FilesetFilterName": ".types",
+ "FilesetMetadataInput": ".types",
+ "FilesetMetadataOutput": ".types",
+ "FilesetOutput": ".types",
+ "FilesetOutputStorage": ".types",
+ "FilesetOutputStorage_Huggingface": ".types",
+ "FilesetOutputStorage_Local": ".types",
+ "FilesetOutputStorage_Ngc": ".types",
+ "FilesetOutputStorage_S3": ".types",
+ "FilesetOutputsPage": ".types",
+ "FilesetPurpose": ".types",
+ "FinetuningType": ".types",
+ "FinetuningTypeFilter": ".types",
+ "FloatFilter": ".types",
+ "Function": ".types",
+ "FunctionCall": ".types",
+ "GLiNerDetection": ".types",
+ "GLiNerDetectionOptions": ".types",
+ "GenerationLog": ".types",
+ "GenerationLogOptions": ".types",
+ "GenerationOptions": ".types",
+ "GenerationOptionsOutputVars": ".types",
+ "GenerationRailsOptions": ".types",
+ "GenerationRailsOptionsInput": ".types",
+ "GenerationRailsOptionsOutput": ".types",
+ "GenerationRailsOptionsRetrieval": ".types",
+ "GenerationStats": ".types",
+ "GenericSortField": ".types",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem": ".jobs",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Docker": ".jobs",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_E2E": ".jobs",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_KubernetesJob": ".jobs",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Subprocess": ".jobs",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_VolcanoJob": ".jobs",
+ "GetTraceApisIntakeV2WorkspacesWorkspaceTracesIdGetRequestMode": ".traces",
+ "GpuExecutionProviderInput": ".types",
+ "GpuExecutionProviderOutput": ".types",
+ "GuardrailCheckRequestFunctionCall": ".guardrails",
+ "GuardrailCheckRequestMessagesItem": ".guardrails",
+ "GuardrailCheckRequestMessagesItem_Assistant": ".guardrails",
+ "GuardrailCheckRequestMessagesItem_Function": ".guardrails",
+ "GuardrailCheckRequestMessagesItem_System": ".guardrails",
+ "GuardrailCheckRequestMessagesItem_Tool": ".guardrails",
+ "GuardrailCheckRequestMessagesItem_User": ".guardrails",
+ "GuardrailCheckRequestStop": ".guardrails",
+ "GuardrailCheckRequestToolChoice": ".guardrails",
+ "GuardrailCheckResponse": ".types",
+ "GuardrailConfig": ".types",
+ "GuardrailConfigData": ".types",
+ "GuardrailConfigFilter": ".types",
+ "GuardrailConfigFilterDescription": ".types",
+ "GuardrailConfigFilterName": ".types",
+ "GuardrailConfigsPage": ".types",
+ "GuardrailsAiRailConfig": ".types",
+ "GuardrailsAiValidatorConfig": ".types",
+ "GuardrailsDataInput": ".types",
+ "GuardrailsDataInputConfig": ".types",
+ "GuardrailsDataOutput": ".types",
+ "HttpValidationError": ".types",
+ "HuggingfaceStorageConfig": ".types",
+ "HuggingfaceStorageConfigRepoType": ".types",
+ "ImagePullSecret": ".types",
+ "ImageUrl": ".types",
+ "ImageUrlDetail": ".types",
+ "InferenceParams": ".types",
+ "IngestResponse": ".types",
+ "InjectionDetection": ".types",
+ "InputRails": ".types",
+ "Instruction": ".types",
+ "InternalServerError": ".errors",
+ "JailbreakDetectionConfig": ".types",
+ "JobExecutionProfileConfig": ".types",
+ "K8SNimOperatorConfig": ".types",
+ "KubernetesEmptyDirVolume": ".types",
+ "KubernetesJobExecutionProfile": ".types",
+ "KubernetesJobExecutionProfileConfig": ".types",
+ "KubernetesJobStorageConfig": ".types",
+ "KubernetesObjectMetadata": ".types",
+ "KubernetesPersistentVolumeClaim": ".types",
+ "KubernetesVolume": ".types",
+ "KubernetesVolumeMount": ".types",
+ "LabelAnnotation": ".types",
+ "LabelAnnotationInput": ".types",
+ "LabelAnnotationInputValue": ".types",
+ "LabelAnnotationInputValueType": ".types",
+ "LabelAnnotationValue": ".types",
+ "LabelAnnotationValueType": ".types",
+ "LinearLayerSpec": ".types",
+ "ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort": ".experiment_groups",
+ "ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort": ".experiments",
+ "ListFilesetFilesResponse": ".types",
+ "ListSpansApisIntakeV2WorkspacesWorkspaceSpansGetRequestMode": ".spans",
+ "ListTracesApisIntakeV2WorkspacesWorkspaceTracesGetRequestMode": ".traces",
+ "LlmCallInfo": ".types",
+ "LocalStorageConfig": ".types",
+ "LogAdapterConfig": ".types",
+ "Lora": ".types",
+ "MambaConfig": ".types",
+ "MessageTemplate": ".types",
+ "MetadataAnnotation": ".types",
+ "MetadataAnnotationInput": ".types",
+ "MiddlewareCall": ".types",
+ "MoEConfig": ".types",
+ "Model": ".types",
+ "ModelCacheConfig": ".types",
+ "ModelDeployment": ".types",
+ "ModelDeploymentConfig": ".types",
+ "ModelDeploymentConfigFilter": ".types",
+ "ModelDeploymentConfigFilterDescription": ".types",
+ "ModelDeploymentConfigFilterName": ".types",
+ "ModelDeploymentConfigModelSpec": ".types",
+ "ModelDeploymentConfigsPage": ".types",
+ "ModelDeploymentFilter": ".types",
+ "ModelDeploymentFilterConfig": ".types",
+ "ModelDeploymentFilterName": ".types",
+ "ModelDeploymentFilterStatusMessage": ".types",
+ "ModelDeploymentStatus": ".types",
+ "ModelDeploymentStatusHistoryItem": ".types",
+ "ModelDeploymentsPage": ".types",
+ "ModelEntity": ".types",
+ "ModelEntityFilter": ".types",
+ "ModelEntityFilterAdapters": ".types",
+ "ModelEntityFilterBaseModel": ".types",
+ "ModelEntityFilterDescription": ".types",
+ "ModelEntityFilterFinetuningType": ".types",
+ "ModelEntityFilterName": ".types",
+ "ModelEntitySortField": ".types",
+ "ModelEntitysPage": ".types",
+ "ModelMetadataContent": ".types",
+ "ModelMode": ".types",
+ "ModelParameters": ".types",
+ "ModelProvider": ".types",
+ "ModelProviderFilter": ".types",
+ "ModelProviderFilterDescription": ".types",
+ "ModelProviderFilterHostUrl": ".types",
+ "ModelProviderFilterName": ".types",
+ "ModelProviderSort": ".types",
+ "ModelProviderStatus": ".types",
+ "ModelProvidersPage": ".types",
+ "ModelSpec": ".types",
+ "ModelType": ".types",
+ "MultilingualConfig": ".types",
+ "NgcStorageConfig": ".types",
+ "NgcStorageConfigTargetType": ".types",
+ "NotFoundError": ".errors",
+ "NoteAnnotation": ".types",
+ "NoteAnnotationInput": ".types",
+ "NumericFilter": ".types",
+ "NvidiaApi": ".client",
+ "OidcDiscoveryResponse": ".types",
+ "OpenAiListModelsResp": ".types",
+ "OpenAiModelResp": ".types",
+ "OtelExportLogsPartialSuccess": ".types",
+ "OtelExportLogsServiceResponse": ".types",
+ "OutputRails": ".types",
+ "OutputRailsStreamingConfig": ".types",
+ "PaginationData": ".types",
+ "PangeaRailConfig": ".types",
+ "PangeaRailOptions": ".types",
+ "PatronusEvaluateApiParams": ".types",
+ "PatronusEvaluateConfigInput": ".types",
+ "PatronusEvaluateConfigOutput": ".types",
+ "PatronusEvaluationSuccessStrategy": ".types",
+ "PatronusRailConfigInput": ".types",
+ "PatronusRailConfigOutput": ".types",
+ "PlatformJobEnvironmentVariable": ".types",
+ "PlatformJobListResultResponse": ".types",
+ "PlatformJobListTaskResponse": ".types",
+ "PlatformJobLog": ".types",
+ "PlatformJobLogPage": ".types",
+ "PlatformJobResponse": ".types",
+ "PlatformJobResponsesPage": ".types",
+ "PlatformJobResultResponse": ".types",
+ "PlatformJobSecretEnvironmentVariableRef": ".types",
+ "PlatformJobSortField": ".types",
+ "PlatformJobSpecInput": ".types",
+ "PlatformJobSpecOutput": ".types",
+ "PlatformJobStatus": ".types",
+ "PlatformJobStatusResponse": ".types",
+ "PlatformJobStep": ".types",
+ "PlatformJobStepSpecInput": ".types",
+ "PlatformJobStepSpecInputExecutor": ".types",
+ "PlatformJobStepSpecInputExecutor_Cpu": ".types",
+ "PlatformJobStepSpecInputExecutor_Gpu": ".types",
+ "PlatformJobStepSpecInputExecutor_GpuDistributed": ".types",
+ "PlatformJobStepSpecInputExecutor_Subprocess": ".types",
+ "PlatformJobStepSpecOutput": ".types",
+ "PlatformJobStepSpecOutputExecutor": ".types",
+ "PlatformJobStepSpecOutputExecutor_Cpu": ".types",
+ "PlatformJobStepSpecOutputExecutor_Gpu": ".types",
+ "PlatformJobStepSpecOutputExecutor_GpuDistributed": ".types",
+ "PlatformJobStepSpecOutputExecutor_Subprocess": ".types",
+ "PlatformJobStepStatusResponse": ".types",
+ "PlatformJobStepWithContext": ".types",
+ "PlatformJobStepWithContextsPage": ".types",
+ "PlatformJobStepsListFilter": ".types",
+ "PlatformJobTask": ".types",
+ "PlatformJobTaskStatusResponse": ".types",
+ "PlatformJobsListFilter": ".types",
+ "PlatformJobsListFilterName": ".types",
+ "PlatformJobsListFilterSource": ".types",
+ "PlatformJobsListFilterStatus": ".types",
+ "PlatformSecretAccessResponse": ".types",
+ "PlatformSecretAdminRotationResponse": ".types",
+ "PlatformSecretResponse": ".types",
+ "PlatformSecretResponsesPage": ".types",
+ "PrivateAiDetection": ".types",
+ "PrivateAiDetectionOptions": ".types",
+ "Project": ".types",
+ "ProjectSortField": ".types",
+ "ProjectsPage": ".types",
+ "PromptData": ".types",
+ "RailStatus": ".types",
+ "RailsConfigDataInput": ".types",
+ "RailsConfigDataOutput": ".types",
+ "RailsConfigInput": ".types",
+ "RailsConfigOutput": ".types",
+ "RailsInput": ".types",
+ "RailsOutput": ".types",
+ "ReasoningConfig": ".types",
+ "RegexDetection": ".types",
+ "RegexDetectionOptions": ".types",
+ "RetrievalRails": ".types",
+ "RoleBinding": ".types",
+ "RoleBindingFilter": ".types",
+ "RoleBindingFilterGrantedBy": ".types",
+ "RoleBindingFilterPrincipal": ".types",
+ "RoleBindingFilterRole": ".types",
+ "RoleBindingsPage": ".types",
+ "S3StorageConfig": ".types",
+ "S3StorageConfigSignatureVersion": ".types",
+ "SecretRef": ".types",
+ "SensitiveDataDetection": ".types",
+ "SensitiveDataDetectionOptions": ".types",
+ "ServedModelMapping": ".types",
+ "ServiceUnavailableError": ".errors",
+ "SingleCallConfig": ".types",
+ "SlidingWindowConfig": ".types",
+ "Span": ".types",
+ "SpanEvaluationContext": ".types",
+ "SpanFilter": ".types",
+ "SpanKind": ".types",
+ "SpanSortField": ".types",
+ "SpanStatus": ".types",
+ "SpansPage": ".types",
+ "StatusEnum": ".types",
+ "StepLifecycle": ".types",
+ "StorageConfigType": ".types",
+ "StringFilter": ".types",
+ "SubprocessExecutionProvider": ".types",
+ "SubprocessJobExecutionProfile": ".types",
+ "SubprocessJobExecutionProfileConfig": ".types",
+ "SubprocessJobExecutionProfileProvider": ".types",
+ "TaskPrompt": ".types",
+ "TaskPromptMessagesItem": ".types",
+ "ToolCallConfig": ".types",
+ "ToolCallingMetadataContent": ".types",
+ "ToolInputRails": ".types",
+ "ToolOutputRails": ".types",
+ "Trace": ".types",
+ "TraceFilter": ".types",
+ "TraceSortField": ".types",
+ "TracesPage": ".types",
+ "TracingConfig": ".types",
+ "TrendMicroRailConfig": ".types",
+ "UnprocessableEntityError": ".errors",
+ "UpdateAdapterRequest": ".types",
+ "UserMessagesConfig": ".types",
+ "ValidationError": ".types",
+ "ValidationErrorLocItem": ".types",
+ "VirtualModel": ".types",
+ "VirtualModelInferenceConfig": ".types",
+ "VirtualModelsPage": ".types",
+ "VolcanoJobExecutionProfile": ".types",
+ "VolcanoJobExecutionProfileConfig": ".types",
+ "Workspace": ".types",
+ "WorkspaceMember": ".types",
+ "WorkspaceMemberListResponse": ".types",
+ "WorkspacesPage": ".types",
+ "adapters": ".adapters",
+ "annotations": ".annotations",
+ "discovery": ".discovery",
+ "entity_store": ".entity_store",
+ "evaluator_results": ".evaluator_results",
+ "experiment_groups": ".experiment_groups",
+ "experiments": ".experiments",
+ "files": ".files",
+ "guardrails": ".guardrails",
+ "iam": ".iam",
+ "inference_gateway": ".inference_gateway",
+ "ingest": ".ingest",
+ "jobs": ".jobs",
+ "model_deployment_configs": ".model_deployment_configs",
+ "model_deployments": ".model_deployments",
+ "model_providers": ".model_providers",
+ "models": ".models",
+ "otlp": ".otlp",
+ "secrets": ".secrets",
+ "secrets_admin": ".secrets_admin",
+ "spans": ".spans",
+ "traces": ".traces",
+ "virtual_models": ".virtual_models",
+}
+
+
+def __getattr__(attr_name: str) -> typing.Any:
+ module_name = _dynamic_imports.get(attr_name)
+ if module_name is None:
+ raise AttributeError(f"No {attr_name} found in _dynamic_imports for module name -> {__name__}")
+ try:
+ module = import_module(module_name, __package__)
+ if module_name == f".{attr_name}":
+ return module
+ else:
+ return getattr(module, attr_name)
+ except ImportError as e:
+ raise ImportError(f"Failed to import {attr_name} from {module_name}: {e}") from e
+ except AttributeError as e:
+ raise AttributeError(f"Failed to get {attr_name} from {module_name}: {e}") from e
+
+
+def __dir__():
+ lazy_attrs = list(_dynamic_imports.keys())
+ return sorted(lazy_attrs)
+
+
+__all__ = [
+ "ActionRails",
+ "ActivatedRail",
+ "Adapter",
+ "AdapterEntityFilter",
+ "AdapterEntityFilterDescription",
+ "AdapterEntityFilterModel",
+ "AdapterEntityFilterName",
+ "AdaptersPage",
+ "AiDefenseRailConfig",
+ "Annotation",
+ "AnnotationFilter",
+ "AnnotationInput",
+ "AnnotationInput_Feedback",
+ "AnnotationInput_Label",
+ "AnnotationInput_Metadata",
+ "AnnotationInput_Note",
+ "AnnotationKind",
+ "AnnotationSortField",
+ "Annotation_Feedback",
+ "Annotation_Label",
+ "Annotation_Metadata",
+ "Annotation_Note",
+ "AnnotationsPage",
+ "ApiEndpointData",
+ "AsyncNvidiaApi",
+ "AtifAgent",
+ "AtifContentPart",
+ "AtifContentPartImage",
+ "AtifContentPartText",
+ "AtifContentPart_Image",
+ "AtifContentPart_Text",
+ "AtifFinalMetrics",
+ "AtifImageSource",
+ "AtifImageSourceMediaType",
+ "AtifIngestRequestSchemaVersion",
+ "AtifMetrics",
+ "AtifObservation",
+ "AtifObservationResult",
+ "AtifObservationResultContent",
+ "AtifStep",
+ "AtifStepAgent",
+ "AtifStepAgentMessage",
+ "AtifStepAgentReasoningEffort",
+ "AtifStepSystem",
+ "AtifStepSystemMessage",
+ "AtifStepUser",
+ "AtifStepUserMessage",
+ "AtifStep_Agent",
+ "AtifStep_System",
+ "AtifStep_User",
+ "AtifSubagentTrajectoryRef",
+ "AtifToolCall",
+ "AuthContext",
+ "AuthDiscoveryResponse",
+ "AutoAlignOptions",
+ "AutoAlignRailConfig",
+ "BackendFormat",
+ "BadRequestError",
+ "BaseModelFilter",
+ "BaseModelFilterName",
+ "CacheStatsConfig",
+ "CacheStatus",
+ "CapturedChatCompletionsRequest",
+ "CapturedChatCompletionsResponse",
+ "CapturedChatMessage",
+ "ChatCompletionAssistantMessageParam",
+ "ChatCompletionContentPartImageParam",
+ "ChatCompletionContentPartTextParam",
+ "ChatCompletionFunctionMessageParam",
+ "ChatCompletionMessageToolCallParam",
+ "ChatCompletionMessageToolCallParamType",
+ "ChatCompletionSystemMessageParam",
+ "ChatCompletionToolMessageParam",
+ "ChatCompletionUserMessageParam",
+ "ChatCompletionUserMessageParamContent",
+ "ChatCompletionUserMessageParamContentOneItem",
+ "ChatCompletionUserMessageParamContentOneItem_ImageUrl",
+ "ChatCompletionUserMessageParamContentOneItem_Text",
+ "ChatCompletionsIngestResponse",
+ "ChatMessageRole",
+ "ClavataRailConfig",
+ "ClavataRailConfigLabelMatchLogic",
+ "ClavataRailOptions",
+ "ComputeResourceSpec",
+ "ComputeResources",
+ "ConflictError",
+ "ContainerExecutorConfig",
+ "ContainerSpec",
+ "ContentSafetyConfig",
+ "CpuExecutionProviderInput",
+ "CpuExecutionProviderOutput",
+ "CreateFilesetRequestStorage",
+ "CreateFilesetRequestStorage_Huggingface",
+ "CreateFilesetRequestStorage_Local",
+ "CreateFilesetRequestStorage_Ngc",
+ "CreateFilesetRequestStorage_S3",
+ "CrowdStrikeAidrRailConfig",
+ "DatasetMetadataContent",
+ "DatasetMetadataContentSchema",
+ "DatasetMetadataContentSchemasByPathValue",
+ "DateRangeFilter",
+ "DatetimeFilter",
+ "DefaultAioHttpClient",
+ "DefaultAsyncHttpxClient",
+ "DeleteResponse",
+ "DialogRails",
+ "DistributedGpuExecutionProviderInput",
+ "DistributedGpuExecutionProviderOutput",
+ "DockerJobExecutionProfile",
+ "DockerJobExecutionProfileConfig",
+ "DockerJobNetworkConfig",
+ "DockerJobStorageConfig",
+ "DockerVolumeMount",
+ "DockerVolumeMountKind",
+ "E2EJobExecutionProfile",
+ "Engine",
+ "EntitiesPage",
+ "Entity",
+ "EvaluationContext",
+ "EvaluatorAggregate",
+ "EvaluatorResult",
+ "EvaluatorResultDataType",
+ "EvaluatorResultFilter",
+ "EvaluatorResultSortField",
+ "EvaluatorResultsPage",
+ "ExecutedAction",
+ "ExperimentContext",
+ "ExperimentFilter",
+ "ExperimentGroupFilter",
+ "ExperimentGroupRequest",
+ "ExperimentGroupResponse",
+ "ExperimentGroupResponsesPage",
+ "ExperimentRequest",
+ "ExperimentResponse",
+ "ExperimentResponsesPage",
+ "ExperimentSessionFilter",
+ "ExperimentSessionResponse",
+ "ExperimentSessionResponsesPage",
+ "FactCheckingRailConfig",
+ "FeedbackAnnotation",
+ "FeedbackAnnotationInput",
+ "FeedbackAnnotationInputValue",
+ "FeedbackAnnotationValue",
+ "FiddlerGuardrails",
+ "FileStorageType",
+ "FilesetFileOutput",
+ "FilesetFilter",
+ "FilesetFilterDescription",
+ "FilesetFilterName",
+ "FilesetMetadataInput",
+ "FilesetMetadataOutput",
+ "FilesetOutput",
+ "FilesetOutputStorage",
+ "FilesetOutputStorage_Huggingface",
+ "FilesetOutputStorage_Local",
+ "FilesetOutputStorage_Ngc",
+ "FilesetOutputStorage_S3",
+ "FilesetOutputsPage",
+ "FilesetPurpose",
+ "FinetuningType",
+ "FinetuningTypeFilter",
+ "FloatFilter",
+ "Function",
+ "FunctionCall",
+ "GLiNerDetection",
+ "GLiNerDetectionOptions",
+ "GenerationLog",
+ "GenerationLogOptions",
+ "GenerationOptions",
+ "GenerationOptionsOutputVars",
+ "GenerationRailsOptions",
+ "GenerationRailsOptionsInput",
+ "GenerationRailsOptionsOutput",
+ "GenerationRailsOptionsRetrieval",
+ "GenerationStats",
+ "GenericSortField",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Docker",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_E2E",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_KubernetesJob",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Subprocess",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_VolcanoJob",
+ "GetTraceApisIntakeV2WorkspacesWorkspaceTracesIdGetRequestMode",
+ "GpuExecutionProviderInput",
+ "GpuExecutionProviderOutput",
+ "GuardrailCheckRequestFunctionCall",
+ "GuardrailCheckRequestMessagesItem",
+ "GuardrailCheckRequestMessagesItem_Assistant",
+ "GuardrailCheckRequestMessagesItem_Function",
+ "GuardrailCheckRequestMessagesItem_System",
+ "GuardrailCheckRequestMessagesItem_Tool",
+ "GuardrailCheckRequestMessagesItem_User",
+ "GuardrailCheckRequestStop",
+ "GuardrailCheckRequestToolChoice",
+ "GuardrailCheckResponse",
+ "GuardrailConfig",
+ "GuardrailConfigData",
+ "GuardrailConfigFilter",
+ "GuardrailConfigFilterDescription",
+ "GuardrailConfigFilterName",
+ "GuardrailConfigsPage",
+ "GuardrailsAiRailConfig",
+ "GuardrailsAiValidatorConfig",
+ "GuardrailsDataInput",
+ "GuardrailsDataInputConfig",
+ "GuardrailsDataOutput",
+ "HttpValidationError",
+ "HuggingfaceStorageConfig",
+ "HuggingfaceStorageConfigRepoType",
+ "ImagePullSecret",
+ "ImageUrl",
+ "ImageUrlDetail",
+ "InferenceParams",
+ "IngestResponse",
+ "InjectionDetection",
+ "InputRails",
+ "Instruction",
+ "InternalServerError",
+ "JailbreakDetectionConfig",
+ "JobExecutionProfileConfig",
+ "K8SNimOperatorConfig",
+ "KubernetesEmptyDirVolume",
+ "KubernetesJobExecutionProfile",
+ "KubernetesJobExecutionProfileConfig",
+ "KubernetesJobStorageConfig",
+ "KubernetesObjectMetadata",
+ "KubernetesPersistentVolumeClaim",
+ "KubernetesVolume",
+ "KubernetesVolumeMount",
+ "LabelAnnotation",
+ "LabelAnnotationInput",
+ "LabelAnnotationInputValue",
+ "LabelAnnotationInputValueType",
+ "LabelAnnotationValue",
+ "LabelAnnotationValueType",
+ "LinearLayerSpec",
+ "ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort",
+ "ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort",
+ "ListFilesetFilesResponse",
+ "ListSpansApisIntakeV2WorkspacesWorkspaceSpansGetRequestMode",
+ "ListTracesApisIntakeV2WorkspacesWorkspaceTracesGetRequestMode",
+ "LlmCallInfo",
+ "LocalStorageConfig",
+ "LogAdapterConfig",
+ "Lora",
+ "MambaConfig",
+ "MessageTemplate",
+ "MetadataAnnotation",
+ "MetadataAnnotationInput",
+ "MiddlewareCall",
+ "MoEConfig",
+ "Model",
+ "ModelCacheConfig",
+ "ModelDeployment",
+ "ModelDeploymentConfig",
+ "ModelDeploymentConfigFilter",
+ "ModelDeploymentConfigFilterDescription",
+ "ModelDeploymentConfigFilterName",
+ "ModelDeploymentConfigModelSpec",
+ "ModelDeploymentConfigsPage",
+ "ModelDeploymentFilter",
+ "ModelDeploymentFilterConfig",
+ "ModelDeploymentFilterName",
+ "ModelDeploymentFilterStatusMessage",
+ "ModelDeploymentStatus",
+ "ModelDeploymentStatusHistoryItem",
+ "ModelDeploymentsPage",
+ "ModelEntity",
+ "ModelEntityFilter",
+ "ModelEntityFilterAdapters",
+ "ModelEntityFilterBaseModel",
+ "ModelEntityFilterDescription",
+ "ModelEntityFilterFinetuningType",
+ "ModelEntityFilterName",
+ "ModelEntitySortField",
+ "ModelEntitysPage",
+ "ModelMetadataContent",
+ "ModelMode",
+ "ModelParameters",
+ "ModelProvider",
+ "ModelProviderFilter",
+ "ModelProviderFilterDescription",
+ "ModelProviderFilterHostUrl",
+ "ModelProviderFilterName",
+ "ModelProviderSort",
+ "ModelProviderStatus",
+ "ModelProvidersPage",
+ "ModelSpec",
+ "ModelType",
+ "MultilingualConfig",
+ "NgcStorageConfig",
+ "NgcStorageConfigTargetType",
+ "NotFoundError",
+ "NoteAnnotation",
+ "NoteAnnotationInput",
+ "NumericFilter",
+ "NvidiaApi",
+ "OidcDiscoveryResponse",
+ "OpenAiListModelsResp",
+ "OpenAiModelResp",
+ "OtelExportLogsPartialSuccess",
+ "OtelExportLogsServiceResponse",
+ "OutputRails",
+ "OutputRailsStreamingConfig",
+ "PaginationData",
+ "PangeaRailConfig",
+ "PangeaRailOptions",
+ "PatronusEvaluateApiParams",
+ "PatronusEvaluateConfigInput",
+ "PatronusEvaluateConfigOutput",
+ "PatronusEvaluationSuccessStrategy",
+ "PatronusRailConfigInput",
+ "PatronusRailConfigOutput",
+ "PlatformJobEnvironmentVariable",
+ "PlatformJobListResultResponse",
+ "PlatformJobListTaskResponse",
+ "PlatformJobLog",
+ "PlatformJobLogPage",
+ "PlatformJobResponse",
+ "PlatformJobResponsesPage",
+ "PlatformJobResultResponse",
+ "PlatformJobSecretEnvironmentVariableRef",
+ "PlatformJobSortField",
+ "PlatformJobSpecInput",
+ "PlatformJobSpecOutput",
+ "PlatformJobStatus",
+ "PlatformJobStatusResponse",
+ "PlatformJobStep",
+ "PlatformJobStepSpecInput",
+ "PlatformJobStepSpecInputExecutor",
+ "PlatformJobStepSpecInputExecutor_Cpu",
+ "PlatformJobStepSpecInputExecutor_Gpu",
+ "PlatformJobStepSpecInputExecutor_GpuDistributed",
+ "PlatformJobStepSpecInputExecutor_Subprocess",
+ "PlatformJobStepSpecOutput",
+ "PlatformJobStepSpecOutputExecutor",
+ "PlatformJobStepSpecOutputExecutor_Cpu",
+ "PlatformJobStepSpecOutputExecutor_Gpu",
+ "PlatformJobStepSpecOutputExecutor_GpuDistributed",
+ "PlatformJobStepSpecOutputExecutor_Subprocess",
+ "PlatformJobStepStatusResponse",
+ "PlatformJobStepWithContext",
+ "PlatformJobStepWithContextsPage",
+ "PlatformJobStepsListFilter",
+ "PlatformJobTask",
+ "PlatformJobTaskStatusResponse",
+ "PlatformJobsListFilter",
+ "PlatformJobsListFilterName",
+ "PlatformJobsListFilterSource",
+ "PlatformJobsListFilterStatus",
+ "PlatformSecretAccessResponse",
+ "PlatformSecretAdminRotationResponse",
+ "PlatformSecretResponse",
+ "PlatformSecretResponsesPage",
+ "PrivateAiDetection",
+ "PrivateAiDetectionOptions",
+ "Project",
+ "ProjectSortField",
+ "ProjectsPage",
+ "PromptData",
+ "RailStatus",
+ "RailsConfigDataInput",
+ "RailsConfigDataOutput",
+ "RailsConfigInput",
+ "RailsConfigOutput",
+ "RailsInput",
+ "RailsOutput",
+ "ReasoningConfig",
+ "RegexDetection",
+ "RegexDetectionOptions",
+ "RetrievalRails",
+ "RoleBinding",
+ "RoleBindingFilter",
+ "RoleBindingFilterGrantedBy",
+ "RoleBindingFilterPrincipal",
+ "RoleBindingFilterRole",
+ "RoleBindingsPage",
+ "S3StorageConfig",
+ "S3StorageConfigSignatureVersion",
+ "SecretRef",
+ "SensitiveDataDetection",
+ "SensitiveDataDetectionOptions",
+ "ServedModelMapping",
+ "ServiceUnavailableError",
+ "SingleCallConfig",
+ "SlidingWindowConfig",
+ "Span",
+ "SpanEvaluationContext",
+ "SpanFilter",
+ "SpanKind",
+ "SpanSortField",
+ "SpanStatus",
+ "SpansPage",
+ "StatusEnum",
+ "StepLifecycle",
+ "StorageConfigType",
+ "StringFilter",
+ "SubprocessExecutionProvider",
+ "SubprocessJobExecutionProfile",
+ "SubprocessJobExecutionProfileConfig",
+ "SubprocessJobExecutionProfileProvider",
+ "TaskPrompt",
+ "TaskPromptMessagesItem",
+ "ToolCallConfig",
+ "ToolCallingMetadataContent",
+ "ToolInputRails",
+ "ToolOutputRails",
+ "Trace",
+ "TraceFilter",
+ "TraceSortField",
+ "TracesPage",
+ "TracingConfig",
+ "TrendMicroRailConfig",
+ "UnprocessableEntityError",
+ "UpdateAdapterRequest",
+ "UserMessagesConfig",
+ "ValidationError",
+ "ValidationErrorLocItem",
+ "VirtualModel",
+ "VirtualModelInferenceConfig",
+ "VirtualModelsPage",
+ "VolcanoJobExecutionProfile",
+ "VolcanoJobExecutionProfileConfig",
+ "Workspace",
+ "WorkspaceMember",
+ "WorkspaceMemberListResponse",
+ "WorkspacesPage",
+ "adapters",
+ "annotations",
+ "discovery",
+ "entity_store",
+ "evaluator_results",
+ "experiment_groups",
+ "experiments",
+ "files",
+ "guardrails",
+ "iam",
+ "inference_gateway",
+ "ingest",
+ "jobs",
+ "model_deployment_configs",
+ "model_deployments",
+ "model_providers",
+ "models",
+ "otlp",
+ "secrets",
+ "secrets_admin",
+ "spans",
+ "traces",
+ "virtual_models",
+]
diff --git a/sdks/python/_default_clients.py b/sdks/python/_default_clients.py
new file mode 100644
index 0000000000..a11056075f
--- /dev/null
+++ b/sdks/python/_default_clients.py
@@ -0,0 +1,30 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+import httpx
+
+SDK_DEFAULT_TIMEOUT = 60
+
+try:
+ import httpx_aiohttp # type: ignore[import-not-found]
+except ImportError:
+
+ class DefaultAioHttpClient(httpx.AsyncClient): # type: ignore
+ def __init__(self, **kwargs: typing.Any) -> None:
+ raise RuntimeError("To use the aiohttp client, install the aiohttp extra: pip install nvidia[aiohttp]")
+
+else:
+
+ class DefaultAioHttpClient(httpx_aiohttp.HttpxAiohttpClient): # type: ignore
+ def __init__(self, **kwargs: typing.Any) -> None:
+ kwargs.setdefault("timeout", SDK_DEFAULT_TIMEOUT)
+ kwargs.setdefault("follow_redirects", True)
+ super().__init__(**kwargs)
+
+
+class DefaultAsyncHttpxClient(httpx.AsyncClient):
+ def __init__(self, **kwargs: typing.Any) -> None:
+ kwargs.setdefault("timeout", SDK_DEFAULT_TIMEOUT)
+ kwargs.setdefault("follow_redirects", True)
+ super().__init__(**kwargs)
diff --git a/sdks/python/adapters/__init__.py b/sdks/python/adapters/__init__.py
new file mode 100644
index 0000000000..5cde0202dc
--- /dev/null
+++ b/sdks/python/adapters/__init__.py
@@ -0,0 +1,4 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
diff --git a/sdks/python/adapters/client.py b/sdks/python/adapters/client.py
new file mode 100644
index 0000000000..d8669e6822
--- /dev/null
+++ b/sdks/python/adapters/client.py
@@ -0,0 +1,577 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.request_options import RequestOptions
+from ..types.adapter import Adapter
+from ..types.adapter_entity_filter import AdapterEntityFilter
+from ..types.adapters_page import AdaptersPage
+from ..types.finetuning_type import FinetuningType
+from ..types.lora import Lora
+from .raw_client import AsyncRawAdaptersClient, RawAdaptersClient
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class AdaptersClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._raw_client = RawAdaptersClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> RawAdaptersClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ RawAdaptersClient
+ """
+ return self._raw_client
+
+ def list_adapters(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[str] = None,
+ filter: typing.Optional[AdapterEntityFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AdaptersPage:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[str]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[AdapterEntityFilter]
+ Filter adapters by name, model (parent model ref string, stored on the adapter), description, fileset, finetuning_type, enabled, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AdaptersPage
+ List adapters in the workspace
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.adapters.list_adapters(
+ workspace="workspace",
+ )
+ """
+ _response = self._raw_client.list_adapters(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ def create_adapter(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ fileset: str,
+ finetuning_type: FinetuningType,
+ model: str,
+ description: typing.Optional[str] = OMIT,
+ enabled: typing.Optional[bool] = OMIT,
+ lora_config: typing.Optional[Lora] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Adapter:
+ """
+ Create an adapter under a base model specified by the "model" field in the body.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Name of the adapter. Name must be unique in the workspace. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ fileset : str
+ Location where adapter files are stored - expected format {workspace}/{fileset_name}
+
+ finetuning_type : FinetuningType
+ Type of finetuning (LORA, P_TUNING, etc.)
+
+ model : str
+ Base model entity.
+ Use `{workspace}/{model_name}` to reference a model in any workspace, or a single `{model_name}` resolved in the path workspace. A single name (2-63 characters) or 'workspace/model_name' where each segment is a valid name (lowercase, digits, hyphens, and temporarily @ . + _; no leading/trailing or consecutive hyphens). If one slash, both sides must be non-empty.
+
+ description : typing.Optional[str]
+ Optional description of the adapter
+
+ enabled : typing.Optional[bool]
+ Whether to make this adapter available for inference post training
+
+ lora_config : typing.Optional[Lora]
+ Lora configuration specifics
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Adapter
+ Create a new adapter for a model
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.adapters.create_adapter(
+ workspace="workspace",
+ name="lora-adapter-v1",
+ fileset="fileset",
+ finetuning_type="lora_merged",
+ model="llama-3-8b-instruct",
+ )
+ """
+ _response = self._raw_client.create_adapter(
+ workspace,
+ name=name,
+ fileset=fileset,
+ finetuning_type=finetuning_type,
+ model=model,
+ description=description,
+ enabled=enabled,
+ lora_config=lora_config,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def get_adapter(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> Adapter:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Adapter
+ Get one adapter by name. Parent model is taken from the adapter's stored parent (entity `parent` field).
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.adapters.get_adapter(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.get_adapter(workspace, name, request_options=request_options)
+ return _response.data
+
+ def delete_adapter(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> None:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.adapters.delete_adapter(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.delete_adapter(workspace, name, request_options=request_options)
+ return _response.data
+
+ def update_adapter(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ enabled: typing.Optional[bool] = OMIT,
+ fileset: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Adapter:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ description : typing.Optional[str]
+ Optional description of the adapter
+
+ enabled : typing.Optional[bool]
+ Whether to make this adapter available for inference post training
+
+ fileset : typing.Optional[str]
+ Updated fileset for the adapter
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Adapter
+ Update adapter metadata. Updates are applied to the row identified by the adapter's stored parent (entity `parent` field).
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.adapters.update_adapter(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.update_adapter(
+ workspace, name, description=description, enabled=enabled, fileset=fileset, request_options=request_options
+ )
+ return _response.data
+
+
+class AsyncAdaptersClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._raw_client = AsyncRawAdaptersClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> AsyncRawAdaptersClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ AsyncRawAdaptersClient
+ """
+ return self._raw_client
+
+ async def list_adapters(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[str] = None,
+ filter: typing.Optional[AdapterEntityFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AdaptersPage:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[str]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[AdapterEntityFilter]
+ Filter adapters by name, model (parent model ref string, stored on the adapter), description, fileset, finetuning_type, enabled, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AdaptersPage
+ List adapters in the workspace
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.adapters.list_adapters(
+ workspace="workspace",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_adapters(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ async def create_adapter(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ fileset: str,
+ finetuning_type: FinetuningType,
+ model: str,
+ description: typing.Optional[str] = OMIT,
+ enabled: typing.Optional[bool] = OMIT,
+ lora_config: typing.Optional[Lora] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Adapter:
+ """
+ Create an adapter under a base model specified by the "model" field in the body.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Name of the adapter. Name must be unique in the workspace. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ fileset : str
+ Location where adapter files are stored - expected format {workspace}/{fileset_name}
+
+ finetuning_type : FinetuningType
+ Type of finetuning (LORA, P_TUNING, etc.)
+
+ model : str
+ Base model entity.
+ Use `{workspace}/{model_name}` to reference a model in any workspace, or a single `{model_name}` resolved in the path workspace. A single name (2-63 characters) or 'workspace/model_name' where each segment is a valid name (lowercase, digits, hyphens, and temporarily @ . + _; no leading/trailing or consecutive hyphens). If one slash, both sides must be non-empty.
+
+ description : typing.Optional[str]
+ Optional description of the adapter
+
+ enabled : typing.Optional[bool]
+ Whether to make this adapter available for inference post training
+
+ lora_config : typing.Optional[Lora]
+ Lora configuration specifics
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Adapter
+ Create a new adapter for a model
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.adapters.create_adapter(
+ workspace="workspace",
+ name="lora-adapter-v1",
+ fileset="fileset",
+ finetuning_type="lora_merged",
+ model="llama-3-8b-instruct",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.create_adapter(
+ workspace,
+ name=name,
+ fileset=fileset,
+ finetuning_type=finetuning_type,
+ model=model,
+ description=description,
+ enabled=enabled,
+ lora_config=lora_config,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def get_adapter(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> Adapter:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Adapter
+ Get one adapter by name. Parent model is taken from the adapter's stored parent (entity `parent` field).
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.adapters.get_adapter(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_adapter(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def delete_adapter(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> None:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.adapters.delete_adapter(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.delete_adapter(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def update_adapter(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ enabled: typing.Optional[bool] = OMIT,
+ fileset: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Adapter:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ description : typing.Optional[str]
+ Optional description of the adapter
+
+ enabled : typing.Optional[bool]
+ Whether to make this adapter available for inference post training
+
+ fileset : typing.Optional[str]
+ Updated fileset for the adapter
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Adapter
+ Update adapter metadata. Updates are applied to the row identified by the adapter's stored parent (entity `parent` field).
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.adapters.update_adapter(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.update_adapter(
+ workspace, name, description=description, enabled=enabled, fileset=fileset, request_options=request_options
+ )
+ return _response.data
diff --git a/sdks/python/adapters/raw_client.py b/sdks/python/adapters/raw_client.py
new file mode 100644
index 0000000000..34b5790cab
--- /dev/null
+++ b/sdks/python/adapters/raw_client.py
@@ -0,0 +1,741 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+from json.decoder import JSONDecodeError
+
+from ..core.api_error import ApiError
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.http_response import AsyncHttpResponse, HttpResponse
+from ..core.jsonable_encoder import encode_path_param
+from ..core.parse_error import ParsingError
+from ..core.pydantic_utilities import parse_obj_as
+from ..core.request_options import RequestOptions
+from ..core.serialization import convert_and_respect_annotation_metadata
+from ..errors.unprocessable_entity_error import UnprocessableEntityError
+from ..types.adapter import Adapter
+from ..types.adapter_entity_filter import AdapterEntityFilter
+from ..types.adapters_page import AdaptersPage
+from ..types.finetuning_type import FinetuningType
+from ..types.lora import Lora
+from pydantic import ValidationError
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class RawAdaptersClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ def list_adapters(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[str] = None,
+ filter: typing.Optional[AdapterEntityFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[AdaptersPage]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[str]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[AdapterEntityFilter]
+ Filter adapters by name, model (parent model ref string, stored on the adapter), description, fileset, finetuning_type, enabled, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[AdaptersPage]
+ List adapters in the workspace
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/adapters",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=AdapterEntityFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ AdaptersPage,
+ parse_obj_as(
+ type_=AdaptersPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def create_adapter(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ fileset: str,
+ finetuning_type: FinetuningType,
+ model: str,
+ description: typing.Optional[str] = OMIT,
+ enabled: typing.Optional[bool] = OMIT,
+ lora_config: typing.Optional[Lora] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[Adapter]:
+ """
+ Create an adapter under a base model specified by the "model" field in the body.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Name of the adapter. Name must be unique in the workspace. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ fileset : str
+ Location where adapter files are stored - expected format {workspace}/{fileset_name}
+
+ finetuning_type : FinetuningType
+ Type of finetuning (LORA, P_TUNING, etc.)
+
+ model : str
+ Base model entity.
+ Use `{workspace}/{model_name}` to reference a model in any workspace, or a single `{model_name}` resolved in the path workspace. A single name (2-63 characters) or 'workspace/model_name' where each segment is a valid name (lowercase, digits, hyphens, and temporarily @ . + _; no leading/trailing or consecutive hyphens). If one slash, both sides must be non-empty.
+
+ description : typing.Optional[str]
+ Optional description of the adapter
+
+ enabled : typing.Optional[bool]
+ Whether to make this adapter available for inference post training
+
+ lora_config : typing.Optional[Lora]
+ Lora configuration specifics
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[Adapter]
+ Create a new adapter for a model
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/adapters",
+ method="POST",
+ json={
+ "name": name,
+ "description": description,
+ "fileset": fileset,
+ "finetuning_type": finetuning_type,
+ "enabled": enabled,
+ "lora_config": convert_and_respect_annotation_metadata(
+ object_=lora_config, annotation=Lora, direction="write"
+ ),
+ "model": model,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Adapter,
+ parse_obj_as(
+ type_=Adapter, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_adapter(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[Adapter]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[Adapter]
+ Get one adapter by name. Parent model is taken from the adapter's stored parent (entity `parent` field).
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/adapters/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Adapter,
+ parse_obj_as(
+ type_=Adapter, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def delete_adapter(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[None]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[None]
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/adapters/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return HttpResponse(response=_response, data=None)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def update_adapter(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ enabled: typing.Optional[bool] = OMIT,
+ fileset: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[Adapter]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ description : typing.Optional[str]
+ Optional description of the adapter
+
+ enabled : typing.Optional[bool]
+ Whether to make this adapter available for inference post training
+
+ fileset : typing.Optional[str]
+ Updated fileset for the adapter
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[Adapter]
+ Update adapter metadata. Updates are applied to the row identified by the adapter's stored parent (entity `parent` field).
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/adapters/{encode_path_param(name)}",
+ method="PATCH",
+ json={
+ "description": description,
+ "enabled": enabled,
+ "fileset": fileset,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Adapter,
+ parse_obj_as(
+ type_=Adapter, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+
+class AsyncRawAdaptersClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ async def list_adapters(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[str] = None,
+ filter: typing.Optional[AdapterEntityFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[AdaptersPage]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[str]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[AdapterEntityFilter]
+ Filter adapters by name, model (parent model ref string, stored on the adapter), description, fileset, finetuning_type, enabled, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[AdaptersPage]
+ List adapters in the workspace
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/adapters",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=AdapterEntityFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ AdaptersPage,
+ parse_obj_as(
+ type_=AdaptersPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def create_adapter(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ fileset: str,
+ finetuning_type: FinetuningType,
+ model: str,
+ description: typing.Optional[str] = OMIT,
+ enabled: typing.Optional[bool] = OMIT,
+ lora_config: typing.Optional[Lora] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[Adapter]:
+ """
+ Create an adapter under a base model specified by the "model" field in the body.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Name of the adapter. Name must be unique in the workspace. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ fileset : str
+ Location where adapter files are stored - expected format {workspace}/{fileset_name}
+
+ finetuning_type : FinetuningType
+ Type of finetuning (LORA, P_TUNING, etc.)
+
+ model : str
+ Base model entity.
+ Use `{workspace}/{model_name}` to reference a model in any workspace, or a single `{model_name}` resolved in the path workspace. A single name (2-63 characters) or 'workspace/model_name' where each segment is a valid name (lowercase, digits, hyphens, and temporarily @ . + _; no leading/trailing or consecutive hyphens). If one slash, both sides must be non-empty.
+
+ description : typing.Optional[str]
+ Optional description of the adapter
+
+ enabled : typing.Optional[bool]
+ Whether to make this adapter available for inference post training
+
+ lora_config : typing.Optional[Lora]
+ Lora configuration specifics
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[Adapter]
+ Create a new adapter for a model
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/adapters",
+ method="POST",
+ json={
+ "name": name,
+ "description": description,
+ "fileset": fileset,
+ "finetuning_type": finetuning_type,
+ "enabled": enabled,
+ "lora_config": convert_and_respect_annotation_metadata(
+ object_=lora_config, annotation=Lora, direction="write"
+ ),
+ "model": model,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Adapter,
+ parse_obj_as(
+ type_=Adapter, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_adapter(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[Adapter]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[Adapter]
+ Get one adapter by name. Parent model is taken from the adapter's stored parent (entity `parent` field).
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/adapters/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Adapter,
+ parse_obj_as(
+ type_=Adapter, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def delete_adapter(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[None]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[None]
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/adapters/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return AsyncHttpResponse(response=_response, data=None)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def update_adapter(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ enabled: typing.Optional[bool] = OMIT,
+ fileset: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[Adapter]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ description : typing.Optional[str]
+ Optional description of the adapter
+
+ enabled : typing.Optional[bool]
+ Whether to make this adapter available for inference post training
+
+ fileset : typing.Optional[str]
+ Updated fileset for the adapter
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[Adapter]
+ Update adapter metadata. Updates are applied to the row identified by the adapter's stored parent (entity `parent` field).
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/adapters/{encode_path_param(name)}",
+ method="PATCH",
+ json={
+ "description": description,
+ "enabled": enabled,
+ "fileset": fileset,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Adapter,
+ parse_obj_as(
+ type_=Adapter, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
diff --git a/sdks/python/annotations/__init__.py b/sdks/python/annotations/__init__.py
new file mode 100644
index 0000000000..5cde0202dc
--- /dev/null
+++ b/sdks/python/annotations/__init__.py
@@ -0,0 +1,4 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
diff --git a/sdks/python/annotations/client.py b/sdks/python/annotations/client.py
new file mode 100644
index 0000000000..c760002fba
--- /dev/null
+++ b/sdks/python/annotations/client.py
@@ -0,0 +1,383 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.request_options import RequestOptions
+from ..types.annotation import Annotation
+from ..types.annotation_filter import AnnotationFilter
+from ..types.annotation_input import AnnotationInput
+from ..types.annotation_sort_field import AnnotationSortField
+from ..types.annotations_page import AnnotationsPage
+from .raw_client import AsyncRawAnnotationsClient, RawAnnotationsClient
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class AnnotationsClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._raw_client = RawAnnotationsClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> RawAnnotationsClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ RawAnnotationsClient
+ """
+ return self._raw_client
+
+ def list_annotations(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[AnnotationSortField] = None,
+ filter: typing.Optional[AnnotationFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AnnotationsPage:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[AnnotationSortField]
+
+ filter : typing.Optional[AnnotationFilter]
+ Filter annotations by span_id, session_id, kind, name, created_by, and created_at range.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AnnotationsPage
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.annotations.list_annotations(
+ workspace="workspace",
+ )
+ """
+ _response = self._raw_client.list_annotations(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ def create_annotation(
+ self, workspace: str, *, request: AnnotationInput, request_options: typing.Optional[RequestOptions] = None
+ ) -> Annotation:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ request : AnnotationInput
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Annotation
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import AnnotationInput_Feedback, NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.annotations.create_annotation(
+ workspace="workspace",
+ request=AnnotationInput_Feedback(
+ session_id="session_id",
+ value="positive",
+ ),
+ )
+ """
+ _response = self._raw_client.create_annotation(workspace, request=request, request_options=request_options)
+ return _response.data
+
+ def get_annotation(
+ self, workspace: str, annotation_id: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> Annotation:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ annotation_id : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Annotation
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.annotations.get_annotation(
+ workspace="workspace",
+ annotation_id="annotation_id",
+ )
+ """
+ _response = self._raw_client.get_annotation(workspace, annotation_id, request_options=request_options)
+ return _response.data
+
+ def delete_annotation(
+ self, workspace: str, annotation_id: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> None:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ annotation_id : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.annotations.delete_annotation(
+ workspace="workspace",
+ annotation_id="annotation_id",
+ )
+ """
+ _response = self._raw_client.delete_annotation(workspace, annotation_id, request_options=request_options)
+ return _response.data
+
+
+class AsyncAnnotationsClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._raw_client = AsyncRawAnnotationsClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> AsyncRawAnnotationsClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ AsyncRawAnnotationsClient
+ """
+ return self._raw_client
+
+ async def list_annotations(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[AnnotationSortField] = None,
+ filter: typing.Optional[AnnotationFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AnnotationsPage:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[AnnotationSortField]
+
+ filter : typing.Optional[AnnotationFilter]
+ Filter annotations by span_id, session_id, kind, name, created_by, and created_at range.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AnnotationsPage
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.annotations.list_annotations(
+ workspace="workspace",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_annotations(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ async def create_annotation(
+ self, workspace: str, *, request: AnnotationInput, request_options: typing.Optional[RequestOptions] = None
+ ) -> Annotation:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ request : AnnotationInput
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Annotation
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AnnotationInput_Feedback, AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.annotations.create_annotation(
+ workspace="workspace",
+ request=AnnotationInput_Feedback(
+ session_id="session_id",
+ value="positive",
+ ),
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.create_annotation(
+ workspace, request=request, request_options=request_options
+ )
+ return _response.data
+
+ async def get_annotation(
+ self, workspace: str, annotation_id: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> Annotation:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ annotation_id : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Annotation
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.annotations.get_annotation(
+ workspace="workspace",
+ annotation_id="annotation_id",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_annotation(workspace, annotation_id, request_options=request_options)
+ return _response.data
+
+ async def delete_annotation(
+ self, workspace: str, annotation_id: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> None:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ annotation_id : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.annotations.delete_annotation(
+ workspace="workspace",
+ annotation_id="annotation_id",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.delete_annotation(workspace, annotation_id, request_options=request_options)
+ return _response.data
diff --git a/sdks/python/annotations/raw_client.py b/sdks/python/annotations/raw_client.py
new file mode 100644
index 0000000000..add3586131
--- /dev/null
+++ b/sdks/python/annotations/raw_client.py
@@ -0,0 +1,548 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+from json.decoder import JSONDecodeError
+
+from ..core.api_error import ApiError
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.http_response import AsyncHttpResponse, HttpResponse
+from ..core.jsonable_encoder import encode_path_param
+from ..core.parse_error import ParsingError
+from ..core.pydantic_utilities import parse_obj_as
+from ..core.request_options import RequestOptions
+from ..core.serialization import convert_and_respect_annotation_metadata
+from ..errors.not_found_error import NotFoundError
+from ..errors.unprocessable_entity_error import UnprocessableEntityError
+from ..types.annotation import Annotation
+from ..types.annotation_filter import AnnotationFilter
+from ..types.annotation_input import AnnotationInput
+from ..types.annotation_sort_field import AnnotationSortField
+from ..types.annotations_page import AnnotationsPage
+from pydantic import ValidationError
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class RawAnnotationsClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ def list_annotations(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[AnnotationSortField] = None,
+ filter: typing.Optional[AnnotationFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[AnnotationsPage]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[AnnotationSortField]
+
+ filter : typing.Optional[AnnotationFilter]
+ Filter annotations by span_id, session_id, kind, name, created_by, and created_at range.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[AnnotationsPage]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/annotations",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=AnnotationFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ AnnotationsPage,
+ parse_obj_as(
+ type_=AnnotationsPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def create_annotation(
+ self, workspace: str, *, request: AnnotationInput, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[Annotation]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ request : AnnotationInput
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[Annotation]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/annotations",
+ method="POST",
+ json=convert_and_respect_annotation_metadata(
+ object_=request, annotation=AnnotationInput, direction="write"
+ ),
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Annotation,
+ parse_obj_as(
+ type_=Annotation, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_annotation(
+ self, workspace: str, annotation_id: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[Annotation]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ annotation_id : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[Annotation]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/annotations/{encode_path_param(annotation_id)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Annotation,
+ parse_obj_as(
+ type_=Annotation, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def delete_annotation(
+ self, workspace: str, annotation_id: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[None]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ annotation_id : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[None]
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/annotations/{encode_path_param(annotation_id)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return HttpResponse(response=_response, data=None)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+
+class AsyncRawAnnotationsClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ async def list_annotations(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[AnnotationSortField] = None,
+ filter: typing.Optional[AnnotationFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[AnnotationsPage]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[AnnotationSortField]
+
+ filter : typing.Optional[AnnotationFilter]
+ Filter annotations by span_id, session_id, kind, name, created_by, and created_at range.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[AnnotationsPage]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/annotations",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=AnnotationFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ AnnotationsPage,
+ parse_obj_as(
+ type_=AnnotationsPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def create_annotation(
+ self, workspace: str, *, request: AnnotationInput, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[Annotation]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ request : AnnotationInput
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[Annotation]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/annotations",
+ method="POST",
+ json=convert_and_respect_annotation_metadata(
+ object_=request, annotation=AnnotationInput, direction="write"
+ ),
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Annotation,
+ parse_obj_as(
+ type_=Annotation, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_annotation(
+ self, workspace: str, annotation_id: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[Annotation]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ annotation_id : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[Annotation]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/annotations/{encode_path_param(annotation_id)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Annotation,
+ parse_obj_as(
+ type_=Annotation, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def delete_annotation(
+ self, workspace: str, annotation_id: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[None]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ annotation_id : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[None]
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/annotations/{encode_path_param(annotation_id)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return AsyncHttpResponse(response=_response, data=None)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
diff --git a/sdks/python/client.py b/sdks/python/client.py
new file mode 100644
index 0000000000..0606417f55
--- /dev/null
+++ b/sdks/python/client.py
@@ -0,0 +1,594 @@
+# This file was auto-generated by Fern from our API Definition.
+
+from __future__ import annotations
+
+import typing
+
+import httpx
+from .core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from .core.logging import LogConfig, Logger
+
+if typing.TYPE_CHECKING:
+ from .adapters.client import AdaptersClient, AsyncAdaptersClient
+ from .annotations.client import AnnotationsClient, AsyncAnnotationsClient
+ from .discovery.client import AsyncDiscoveryClient, DiscoveryClient
+ from .entity_store.client import AsyncEntityStoreClient, EntityStoreClient
+ from .evaluator_results.client import AsyncEvaluatorResultsClient, EvaluatorResultsClient
+ from .experiment_groups.client import AsyncExperimentGroupsClient, ExperimentGroupsClient
+ from .experiments.client import AsyncExperimentsClient, ExperimentsClient
+ from .files.client import AsyncFilesClient, FilesClient
+ from .guardrails.client import AsyncGuardrailsClient, GuardrailsClient
+ from .iam.client import AsyncIamClient, IamClient
+ from .inference_gateway.client import AsyncInferenceGatewayClient, InferenceGatewayClient
+ from .ingest.client import AsyncIngestClient, IngestClient
+ from .jobs.client import AsyncJobsClient, JobsClient
+ from .model_deployment_configs.client import AsyncModelDeploymentConfigsClient, ModelDeploymentConfigsClient
+ from .model_deployments.client import AsyncModelDeploymentsClient, ModelDeploymentsClient
+ from .model_providers.client import AsyncModelProvidersClient, ModelProvidersClient
+ from .models.client import AsyncModelsClient, ModelsClient
+ from .otlp.client import AsyncOtlpClient, OtlpClient
+ from .secrets.client import AsyncSecretsClient, SecretsClient
+ from .secrets_admin.client import AsyncSecretsAdminClient, SecretsAdminClient
+ from .spans.client import AsyncSpansClient, SpansClient
+ from .traces.client import AsyncTracesClient, TracesClient
+ from .virtual_models.client import AsyncVirtualModelsClient, VirtualModelsClient
+
+
+class NvidiaApi:
+ """
+ Use this class to access the different functions within the SDK. You can instantiate any number of clients with different configuration that will propagate to these functions.
+
+ Parameters
+ ----------
+ base_url : str
+ The base url to use for requests from the client.
+
+ headers : typing.Optional[typing.Dict[str, str]]
+ Additional headers to send with every request.
+
+ timeout : typing.Optional[float]
+ The timeout to be used, in seconds, for requests. By default the timeout is 60 seconds, unless a custom httpx client is used, in which case this default is not enforced.
+
+ max_retries : typing.Optional[int]
+ The default maximum number of retries for failed requests. Defaults to 2. Per-request `max_retries` in `request_options` takes precedence over this value.
+
+ follow_redirects : typing.Optional[bool]
+ Whether the default httpx client follows redirects or not, this is irrelevant if a custom httpx client is passed in.
+
+ httpx_client : typing.Optional[httpx.Client]
+ The httpx client to use for making requests, a preconfigured client is used by default, however this is useful should you want to pass in any custom httpx configuration.
+
+ logging : typing.Optional[typing.Union[LogConfig, Logger]]
+ Configure logging for the SDK. Accepts a LogConfig dict with 'level' (debug/info/warn/error), 'logger' (custom logger implementation), and 'silent' (boolean, defaults to True) fields. You can also pass a pre-configured Logger instance.
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ """
+
+ def __init__(
+ self,
+ *,
+ base_url: str,
+ headers: typing.Optional[typing.Dict[str, str]] = None,
+ timeout: typing.Optional[float] = None,
+ max_retries: typing.Optional[int] = None,
+ follow_redirects: typing.Optional[bool] = True,
+ httpx_client: typing.Optional[httpx.Client] = None,
+ logging: typing.Optional[typing.Union[LogConfig, Logger]] = None,
+ ):
+ _defaulted_timeout = (
+ timeout if timeout is not None else 60 if httpx_client is None else httpx_client.timeout.read
+ )
+ _defaulted_max_retries = max_retries if max_retries is not None else 2
+ self._client_wrapper = SyncClientWrapper(
+ base_url=base_url,
+ headers=headers,
+ httpx_client=httpx_client
+ if httpx_client is not None
+ else httpx.Client(timeout=_defaulted_timeout, follow_redirects=follow_redirects)
+ if follow_redirects is not None
+ else httpx.Client(timeout=_defaulted_timeout),
+ timeout=_defaulted_timeout,
+ max_retries=_defaulted_max_retries,
+ logging=logging,
+ )
+ self._discovery: typing.Optional[DiscoveryClient] = None
+ self._iam: typing.Optional[IamClient] = None
+ self._entity_store: typing.Optional[EntityStoreClient] = None
+ self._files: typing.Optional[FilesClient] = None
+ self._otlp: typing.Optional[OtlpClient] = None
+ self._guardrails: typing.Optional[GuardrailsClient] = None
+ self._inference_gateway: typing.Optional[InferenceGatewayClient] = None
+ self._virtual_models: typing.Optional[VirtualModelsClient] = None
+ self._annotations: typing.Optional[AnnotationsClient] = None
+ self._evaluator_results: typing.Optional[EvaluatorResultsClient] = None
+ self._experiment_groups: typing.Optional[ExperimentGroupsClient] = None
+ self._experiments: typing.Optional[ExperimentsClient] = None
+ self._ingest: typing.Optional[IngestClient] = None
+ self._spans: typing.Optional[SpansClient] = None
+ self._traces: typing.Optional[TracesClient] = None
+ self._jobs: typing.Optional[JobsClient] = None
+ self._adapters: typing.Optional[AdaptersClient] = None
+ self._model_deployment_configs: typing.Optional[ModelDeploymentConfigsClient] = None
+ self._model_deployments: typing.Optional[ModelDeploymentsClient] = None
+ self._models: typing.Optional[ModelsClient] = None
+ self._model_providers: typing.Optional[ModelProvidersClient] = None
+ self._secrets_admin: typing.Optional[SecretsAdminClient] = None
+ self._secrets: typing.Optional[SecretsClient] = None
+
+ @property
+ def discovery(self):
+ if self._discovery is None:
+ from .discovery.client import DiscoveryClient # noqa: E402
+
+ self._discovery = DiscoveryClient(client_wrapper=self._client_wrapper)
+ return self._discovery
+
+ @property
+ def iam(self):
+ if self._iam is None:
+ from .iam.client import IamClient # noqa: E402
+
+ self._iam = IamClient(client_wrapper=self._client_wrapper)
+ return self._iam
+
+ @property
+ def entity_store(self):
+ if self._entity_store is None:
+ from .entity_store.client import EntityStoreClient # noqa: E402
+
+ self._entity_store = EntityStoreClient(client_wrapper=self._client_wrapper)
+ return self._entity_store
+
+ @property
+ def files(self):
+ if self._files is None:
+ from .files.client import FilesClient # noqa: E402
+
+ self._files = FilesClient(client_wrapper=self._client_wrapper)
+ return self._files
+
+ @property
+ def otlp(self):
+ if self._otlp is None:
+ from .otlp.client import OtlpClient # noqa: E402
+
+ self._otlp = OtlpClient(client_wrapper=self._client_wrapper)
+ return self._otlp
+
+ @property
+ def guardrails(self):
+ if self._guardrails is None:
+ from .guardrails.client import GuardrailsClient # noqa: E402
+
+ self._guardrails = GuardrailsClient(client_wrapper=self._client_wrapper)
+ return self._guardrails
+
+ @property
+ def inference_gateway(self):
+ if self._inference_gateway is None:
+ from .inference_gateway.client import InferenceGatewayClient # noqa: E402
+
+ self._inference_gateway = InferenceGatewayClient(client_wrapper=self._client_wrapper)
+ return self._inference_gateway
+
+ @property
+ def virtual_models(self):
+ if self._virtual_models is None:
+ from .virtual_models.client import VirtualModelsClient # noqa: E402
+
+ self._virtual_models = VirtualModelsClient(client_wrapper=self._client_wrapper)
+ return self._virtual_models
+
+ @property
+ def annotations(self):
+ if self._annotations is None:
+ from .annotations.client import AnnotationsClient # noqa: E402
+
+ self._annotations = AnnotationsClient(client_wrapper=self._client_wrapper)
+ return self._annotations
+
+ @property
+ def evaluator_results(self):
+ if self._evaluator_results is None:
+ from .evaluator_results.client import EvaluatorResultsClient # noqa: E402
+
+ self._evaluator_results = EvaluatorResultsClient(client_wrapper=self._client_wrapper)
+ return self._evaluator_results
+
+ @property
+ def experiment_groups(self):
+ if self._experiment_groups is None:
+ from .experiment_groups.client import ExperimentGroupsClient # noqa: E402
+
+ self._experiment_groups = ExperimentGroupsClient(client_wrapper=self._client_wrapper)
+ return self._experiment_groups
+
+ @property
+ def experiments(self):
+ if self._experiments is None:
+ from .experiments.client import ExperimentsClient # noqa: E402
+
+ self._experiments = ExperimentsClient(client_wrapper=self._client_wrapper)
+ return self._experiments
+
+ @property
+ def ingest(self):
+ if self._ingest is None:
+ from .ingest.client import IngestClient # noqa: E402
+
+ self._ingest = IngestClient(client_wrapper=self._client_wrapper)
+ return self._ingest
+
+ @property
+ def spans(self):
+ if self._spans is None:
+ from .spans.client import SpansClient # noqa: E402
+
+ self._spans = SpansClient(client_wrapper=self._client_wrapper)
+ return self._spans
+
+ @property
+ def traces(self):
+ if self._traces is None:
+ from .traces.client import TracesClient # noqa: E402
+
+ self._traces = TracesClient(client_wrapper=self._client_wrapper)
+ return self._traces
+
+ @property
+ def jobs(self):
+ if self._jobs is None:
+ from .jobs.client import JobsClient # noqa: E402
+
+ self._jobs = JobsClient(client_wrapper=self._client_wrapper)
+ return self._jobs
+
+ @property
+ def adapters(self):
+ if self._adapters is None:
+ from .adapters.client import AdaptersClient # noqa: E402
+
+ self._adapters = AdaptersClient(client_wrapper=self._client_wrapper)
+ return self._adapters
+
+ @property
+ def model_deployment_configs(self):
+ if self._model_deployment_configs is None:
+ from .model_deployment_configs.client import ModelDeploymentConfigsClient # noqa: E402
+
+ self._model_deployment_configs = ModelDeploymentConfigsClient(client_wrapper=self._client_wrapper)
+ return self._model_deployment_configs
+
+ @property
+ def model_deployments(self):
+ if self._model_deployments is None:
+ from .model_deployments.client import ModelDeploymentsClient # noqa: E402
+
+ self._model_deployments = ModelDeploymentsClient(client_wrapper=self._client_wrapper)
+ return self._model_deployments
+
+ @property
+ def models(self):
+ if self._models is None:
+ from .models.client import ModelsClient # noqa: E402
+
+ self._models = ModelsClient(client_wrapper=self._client_wrapper)
+ return self._models
+
+ @property
+ def model_providers(self):
+ if self._model_providers is None:
+ from .model_providers.client import ModelProvidersClient # noqa: E402
+
+ self._model_providers = ModelProvidersClient(client_wrapper=self._client_wrapper)
+ return self._model_providers
+
+ @property
+ def secrets_admin(self):
+ if self._secrets_admin is None:
+ from .secrets_admin.client import SecretsAdminClient # noqa: E402
+
+ self._secrets_admin = SecretsAdminClient(client_wrapper=self._client_wrapper)
+ return self._secrets_admin
+
+ @property
+ def secrets(self):
+ if self._secrets is None:
+ from .secrets.client import SecretsClient # noqa: E402
+
+ self._secrets = SecretsClient(client_wrapper=self._client_wrapper)
+ return self._secrets
+
+
+def _make_default_async_client(
+ timeout: typing.Optional[float],
+ follow_redirects: typing.Optional[bool],
+) -> httpx.AsyncClient:
+ try:
+ import httpx_aiohttp # type: ignore[import-not-found]
+ except ImportError:
+ pass
+ else:
+ if follow_redirects is not None:
+ return httpx_aiohttp.HttpxAiohttpClient(timeout=timeout, follow_redirects=follow_redirects)
+ return httpx_aiohttp.HttpxAiohttpClient(timeout=timeout)
+
+ if follow_redirects is not None:
+ return httpx.AsyncClient(timeout=timeout, follow_redirects=follow_redirects)
+ return httpx.AsyncClient(timeout=timeout)
+
+
+class AsyncNvidiaApi:
+ """
+ Use this class to access the different functions within the SDK. You can instantiate any number of clients with different configuration that will propagate to these functions.
+
+ Parameters
+ ----------
+ base_url : str
+ The base url to use for requests from the client.
+
+ headers : typing.Optional[typing.Dict[str, str]]
+ Additional headers to send with every request.
+
+ timeout : typing.Optional[float]
+ The timeout to be used, in seconds, for requests. By default the timeout is 60 seconds, unless a custom httpx client is used, in which case this default is not enforced.
+
+ max_retries : typing.Optional[int]
+ The default maximum number of retries for failed requests. Defaults to 2. Per-request `max_retries` in `request_options` takes precedence over this value.
+
+ follow_redirects : typing.Optional[bool]
+ Whether the default httpx client follows redirects or not, this is irrelevant if a custom httpx client is passed in.
+
+ httpx_client : typing.Optional[httpx.AsyncClient]
+ The httpx client to use for making requests, a preconfigured client is used by default, however this is useful should you want to pass in any custom httpx configuration.
+
+ logging : typing.Optional[typing.Union[LogConfig, Logger]]
+ Configure logging for the SDK. Accepts a LogConfig dict with 'level' (debug/info/warn/error), 'logger' (custom logger implementation), and 'silent' (boolean, defaults to True) fields. You can also pass a pre-configured Logger instance.
+
+ Examples
+ --------
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ """
+
+ def __init__(
+ self,
+ *,
+ base_url: str,
+ headers: typing.Optional[typing.Dict[str, str]] = None,
+ timeout: typing.Optional[float] = None,
+ max_retries: typing.Optional[int] = None,
+ follow_redirects: typing.Optional[bool] = True,
+ httpx_client: typing.Optional[httpx.AsyncClient] = None,
+ logging: typing.Optional[typing.Union[LogConfig, Logger]] = None,
+ ):
+ _defaulted_timeout = (
+ timeout if timeout is not None else 60 if httpx_client is None else httpx_client.timeout.read
+ )
+ _defaulted_max_retries = max_retries if max_retries is not None else 2
+ self._client_wrapper = AsyncClientWrapper(
+ base_url=base_url,
+ headers=headers,
+ httpx_client=httpx_client
+ if httpx_client is not None
+ else _make_default_async_client(timeout=_defaulted_timeout, follow_redirects=follow_redirects),
+ timeout=_defaulted_timeout,
+ max_retries=_defaulted_max_retries,
+ logging=logging,
+ )
+ self._discovery: typing.Optional[AsyncDiscoveryClient] = None
+ self._iam: typing.Optional[AsyncIamClient] = None
+ self._entity_store: typing.Optional[AsyncEntityStoreClient] = None
+ self._files: typing.Optional[AsyncFilesClient] = None
+ self._otlp: typing.Optional[AsyncOtlpClient] = None
+ self._guardrails: typing.Optional[AsyncGuardrailsClient] = None
+ self._inference_gateway: typing.Optional[AsyncInferenceGatewayClient] = None
+ self._virtual_models: typing.Optional[AsyncVirtualModelsClient] = None
+ self._annotations: typing.Optional[AsyncAnnotationsClient] = None
+ self._evaluator_results: typing.Optional[AsyncEvaluatorResultsClient] = None
+ self._experiment_groups: typing.Optional[AsyncExperimentGroupsClient] = None
+ self._experiments: typing.Optional[AsyncExperimentsClient] = None
+ self._ingest: typing.Optional[AsyncIngestClient] = None
+ self._spans: typing.Optional[AsyncSpansClient] = None
+ self._traces: typing.Optional[AsyncTracesClient] = None
+ self._jobs: typing.Optional[AsyncJobsClient] = None
+ self._adapters: typing.Optional[AsyncAdaptersClient] = None
+ self._model_deployment_configs: typing.Optional[AsyncModelDeploymentConfigsClient] = None
+ self._model_deployments: typing.Optional[AsyncModelDeploymentsClient] = None
+ self._models: typing.Optional[AsyncModelsClient] = None
+ self._model_providers: typing.Optional[AsyncModelProvidersClient] = None
+ self._secrets_admin: typing.Optional[AsyncSecretsAdminClient] = None
+ self._secrets: typing.Optional[AsyncSecretsClient] = None
+
+ @property
+ def discovery(self):
+ if self._discovery is None:
+ from .discovery.client import AsyncDiscoveryClient # noqa: E402
+
+ self._discovery = AsyncDiscoveryClient(client_wrapper=self._client_wrapper)
+ return self._discovery
+
+ @property
+ def iam(self):
+ if self._iam is None:
+ from .iam.client import AsyncIamClient # noqa: E402
+
+ self._iam = AsyncIamClient(client_wrapper=self._client_wrapper)
+ return self._iam
+
+ @property
+ def entity_store(self):
+ if self._entity_store is None:
+ from .entity_store.client import AsyncEntityStoreClient # noqa: E402
+
+ self._entity_store = AsyncEntityStoreClient(client_wrapper=self._client_wrapper)
+ return self._entity_store
+
+ @property
+ def files(self):
+ if self._files is None:
+ from .files.client import AsyncFilesClient # noqa: E402
+
+ self._files = AsyncFilesClient(client_wrapper=self._client_wrapper)
+ return self._files
+
+ @property
+ def otlp(self):
+ if self._otlp is None:
+ from .otlp.client import AsyncOtlpClient # noqa: E402
+
+ self._otlp = AsyncOtlpClient(client_wrapper=self._client_wrapper)
+ return self._otlp
+
+ @property
+ def guardrails(self):
+ if self._guardrails is None:
+ from .guardrails.client import AsyncGuardrailsClient # noqa: E402
+
+ self._guardrails = AsyncGuardrailsClient(client_wrapper=self._client_wrapper)
+ return self._guardrails
+
+ @property
+ def inference_gateway(self):
+ if self._inference_gateway is None:
+ from .inference_gateway.client import AsyncInferenceGatewayClient # noqa: E402
+
+ self._inference_gateway = AsyncInferenceGatewayClient(client_wrapper=self._client_wrapper)
+ return self._inference_gateway
+
+ @property
+ def virtual_models(self):
+ if self._virtual_models is None:
+ from .virtual_models.client import AsyncVirtualModelsClient # noqa: E402
+
+ self._virtual_models = AsyncVirtualModelsClient(client_wrapper=self._client_wrapper)
+ return self._virtual_models
+
+ @property
+ def annotations(self):
+ if self._annotations is None:
+ from .annotations.client import AsyncAnnotationsClient # noqa: E402
+
+ self._annotations = AsyncAnnotationsClient(client_wrapper=self._client_wrapper)
+ return self._annotations
+
+ @property
+ def evaluator_results(self):
+ if self._evaluator_results is None:
+ from .evaluator_results.client import AsyncEvaluatorResultsClient # noqa: E402
+
+ self._evaluator_results = AsyncEvaluatorResultsClient(client_wrapper=self._client_wrapper)
+ return self._evaluator_results
+
+ @property
+ def experiment_groups(self):
+ if self._experiment_groups is None:
+ from .experiment_groups.client import AsyncExperimentGroupsClient # noqa: E402
+
+ self._experiment_groups = AsyncExperimentGroupsClient(client_wrapper=self._client_wrapper)
+ return self._experiment_groups
+
+ @property
+ def experiments(self):
+ if self._experiments is None:
+ from .experiments.client import AsyncExperimentsClient # noqa: E402
+
+ self._experiments = AsyncExperimentsClient(client_wrapper=self._client_wrapper)
+ return self._experiments
+
+ @property
+ def ingest(self):
+ if self._ingest is None:
+ from .ingest.client import AsyncIngestClient # noqa: E402
+
+ self._ingest = AsyncIngestClient(client_wrapper=self._client_wrapper)
+ return self._ingest
+
+ @property
+ def spans(self):
+ if self._spans is None:
+ from .spans.client import AsyncSpansClient # noqa: E402
+
+ self._spans = AsyncSpansClient(client_wrapper=self._client_wrapper)
+ return self._spans
+
+ @property
+ def traces(self):
+ if self._traces is None:
+ from .traces.client import AsyncTracesClient # noqa: E402
+
+ self._traces = AsyncTracesClient(client_wrapper=self._client_wrapper)
+ return self._traces
+
+ @property
+ def jobs(self):
+ if self._jobs is None:
+ from .jobs.client import AsyncJobsClient # noqa: E402
+
+ self._jobs = AsyncJobsClient(client_wrapper=self._client_wrapper)
+ return self._jobs
+
+ @property
+ def adapters(self):
+ if self._adapters is None:
+ from .adapters.client import AsyncAdaptersClient # noqa: E402
+
+ self._adapters = AsyncAdaptersClient(client_wrapper=self._client_wrapper)
+ return self._adapters
+
+ @property
+ def model_deployment_configs(self):
+ if self._model_deployment_configs is None:
+ from .model_deployment_configs.client import AsyncModelDeploymentConfigsClient # noqa: E402
+
+ self._model_deployment_configs = AsyncModelDeploymentConfigsClient(client_wrapper=self._client_wrapper)
+ return self._model_deployment_configs
+
+ @property
+ def model_deployments(self):
+ if self._model_deployments is None:
+ from .model_deployments.client import AsyncModelDeploymentsClient # noqa: E402
+
+ self._model_deployments = AsyncModelDeploymentsClient(client_wrapper=self._client_wrapper)
+ return self._model_deployments
+
+ @property
+ def models(self):
+ if self._models is None:
+ from .models.client import AsyncModelsClient # noqa: E402
+
+ self._models = AsyncModelsClient(client_wrapper=self._client_wrapper)
+ return self._models
+
+ @property
+ def model_providers(self):
+ if self._model_providers is None:
+ from .model_providers.client import AsyncModelProvidersClient # noqa: E402
+
+ self._model_providers = AsyncModelProvidersClient(client_wrapper=self._client_wrapper)
+ return self._model_providers
+
+ @property
+ def secrets_admin(self):
+ if self._secrets_admin is None:
+ from .secrets_admin.client import AsyncSecretsAdminClient # noqa: E402
+
+ self._secrets_admin = AsyncSecretsAdminClient(client_wrapper=self._client_wrapper)
+ return self._secrets_admin
+
+ @property
+ def secrets(self):
+ if self._secrets is None:
+ from .secrets.client import AsyncSecretsClient # noqa: E402
+
+ self._secrets = AsyncSecretsClient(client_wrapper=self._client_wrapper)
+ return self._secrets
diff --git a/sdks/python/core/__init__.py b/sdks/python/core/__init__.py
new file mode 100644
index 0000000000..5bc159a110
--- /dev/null
+++ b/sdks/python/core/__init__.py
@@ -0,0 +1,127 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
+import typing
+from importlib import import_module
+
+if typing.TYPE_CHECKING:
+ from .api_error import ApiError
+ from .client_wrapper import AsyncClientWrapper, BaseClientWrapper, SyncClientWrapper
+ from .datetime_utils import Rfc2822DateTime, parse_rfc2822_datetime, serialize_datetime
+ from .file import File, convert_file_dict_to_httpx_tuples, with_content_type
+ from .http_client import AsyncHttpClient, HttpClient
+ from .http_response import AsyncHttpResponse, HttpResponse
+ from .jsonable_encoder import encode_path_param, jsonable_encoder
+ from .logging import ConsoleLogger, ILogger, LogConfig, LogLevel, Logger, create_logger
+ from .parse_error import ParsingError
+ from .pydantic_utilities import (
+ IS_PYDANTIC_V2,
+ UniversalBaseModel,
+ UniversalRootModel,
+ parse_obj_as,
+ universal_field_validator,
+ universal_root_validator,
+ update_forward_refs,
+ )
+ from .query_encoder import encode_query
+ from .remove_none_from_dict import remove_none_from_dict
+ from .request_options import RequestOptions
+ from .serialization import FieldMetadata, convert_and_respect_annotation_metadata
+_dynamic_imports: typing.Dict[str, str] = {
+ "ApiError": ".api_error",
+ "AsyncClientWrapper": ".client_wrapper",
+ "AsyncHttpClient": ".http_client",
+ "AsyncHttpResponse": ".http_response",
+ "BaseClientWrapper": ".client_wrapper",
+ "ConsoleLogger": ".logging",
+ "FieldMetadata": ".serialization",
+ "File": ".file",
+ "HttpClient": ".http_client",
+ "HttpResponse": ".http_response",
+ "ILogger": ".logging",
+ "IS_PYDANTIC_V2": ".pydantic_utilities",
+ "LogConfig": ".logging",
+ "LogLevel": ".logging",
+ "Logger": ".logging",
+ "ParsingError": ".parse_error",
+ "RequestOptions": ".request_options",
+ "Rfc2822DateTime": ".datetime_utils",
+ "SyncClientWrapper": ".client_wrapper",
+ "UniversalBaseModel": ".pydantic_utilities",
+ "UniversalRootModel": ".pydantic_utilities",
+ "convert_and_respect_annotation_metadata": ".serialization",
+ "convert_file_dict_to_httpx_tuples": ".file",
+ "create_logger": ".logging",
+ "encode_path_param": ".jsonable_encoder",
+ "encode_query": ".query_encoder",
+ "jsonable_encoder": ".jsonable_encoder",
+ "parse_obj_as": ".pydantic_utilities",
+ "parse_rfc2822_datetime": ".datetime_utils",
+ "remove_none_from_dict": ".remove_none_from_dict",
+ "serialize_datetime": ".datetime_utils",
+ "universal_field_validator": ".pydantic_utilities",
+ "universal_root_validator": ".pydantic_utilities",
+ "update_forward_refs": ".pydantic_utilities",
+ "with_content_type": ".file",
+}
+
+
+def __getattr__(attr_name: str) -> typing.Any:
+ module_name = _dynamic_imports.get(attr_name)
+ if module_name is None:
+ raise AttributeError(f"No {attr_name} found in _dynamic_imports for module name -> {__name__}")
+ try:
+ module = import_module(module_name, __package__)
+ if module_name == f".{attr_name}":
+ return module
+ else:
+ return getattr(module, attr_name)
+ except ImportError as e:
+ raise ImportError(f"Failed to import {attr_name} from {module_name}: {e}") from e
+ except AttributeError as e:
+ raise AttributeError(f"Failed to get {attr_name} from {module_name}: {e}") from e
+
+
+def __dir__():
+ lazy_attrs = list(_dynamic_imports.keys())
+ return sorted(lazy_attrs)
+
+
+__all__ = [
+ "ApiError",
+ "AsyncClientWrapper",
+ "AsyncHttpClient",
+ "AsyncHttpResponse",
+ "BaseClientWrapper",
+ "ConsoleLogger",
+ "FieldMetadata",
+ "File",
+ "HttpClient",
+ "HttpResponse",
+ "ILogger",
+ "IS_PYDANTIC_V2",
+ "LogConfig",
+ "LogLevel",
+ "Logger",
+ "ParsingError",
+ "RequestOptions",
+ "Rfc2822DateTime",
+ "SyncClientWrapper",
+ "UniversalBaseModel",
+ "UniversalRootModel",
+ "convert_and_respect_annotation_metadata",
+ "convert_file_dict_to_httpx_tuples",
+ "create_logger",
+ "encode_path_param",
+ "encode_query",
+ "jsonable_encoder",
+ "parse_obj_as",
+ "parse_rfc2822_datetime",
+ "remove_none_from_dict",
+ "serialize_datetime",
+ "universal_field_validator",
+ "universal_root_validator",
+ "update_forward_refs",
+ "with_content_type",
+]
diff --git a/sdks/python/core/api_error.py b/sdks/python/core/api_error.py
new file mode 100644
index 0000000000..6f850a60cb
--- /dev/null
+++ b/sdks/python/core/api_error.py
@@ -0,0 +1,23 @@
+# This file was auto-generated by Fern from our API Definition.
+
+from typing import Any, Dict, Optional
+
+
+class ApiError(Exception):
+ headers: Optional[Dict[str, str]]
+ status_code: Optional[int]
+ body: Any
+
+ def __init__(
+ self,
+ *,
+ headers: Optional[Dict[str, str]] = None,
+ status_code: Optional[int] = None,
+ body: Any = None,
+ ) -> None:
+ self.headers = headers
+ self.status_code = status_code
+ self.body = body
+
+ def __str__(self) -> str:
+ return f"headers: {self.headers}, status_code: {self.status_code}, body: {self.body}"
diff --git a/sdks/python/core/client_wrapper.py b/sdks/python/core/client_wrapper.py
new file mode 100644
index 0000000000..217123f9cb
--- /dev/null
+++ b/sdks/python/core/client_wrapper.py
@@ -0,0 +1,101 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+import httpx
+from .http_client import AsyncHttpClient, HttpClient
+from .logging import LogConfig, Logger
+
+
+class BaseClientWrapper:
+ def __init__(
+ self,
+ *,
+ headers: typing.Optional[typing.Dict[str, str]] = None,
+ base_url: str,
+ timeout: typing.Optional[float] = None,
+ max_retries: int = 2,
+ logging: typing.Optional[typing.Union[LogConfig, Logger]] = None,
+ ):
+ self._headers = headers
+ self._base_url = base_url
+ self._timeout = timeout
+ self._max_retries = max_retries
+ self._logging = logging
+
+ def get_headers(self) -> typing.Dict[str, str]:
+ import platform
+
+ headers: typing.Dict[str, str] = {
+ "X-Fern-Language": "Python",
+ "X-Fern-Runtime": f"python/{platform.python_version()}",
+ "X-Fern-Platform": f"{platform.system().lower()}/{platform.release()}",
+ **(self.get_custom_headers() or {}),
+ }
+ return headers
+
+ def get_custom_headers(self) -> typing.Optional[typing.Dict[str, str]]:
+ return self._headers
+
+ def get_base_url(self) -> str:
+ return self._base_url
+
+ def get_timeout(self) -> typing.Optional[float]:
+ return self._timeout
+
+ def get_max_retries(self) -> int:
+ return self._max_retries
+
+
+class SyncClientWrapper(BaseClientWrapper):
+ def __init__(
+ self,
+ *,
+ headers: typing.Optional[typing.Dict[str, str]] = None,
+ base_url: str,
+ timeout: typing.Optional[float] = None,
+ max_retries: int = 2,
+ logging: typing.Optional[typing.Union[LogConfig, Logger]] = None,
+ httpx_client: httpx.Client,
+ ):
+ super().__init__(headers=headers, base_url=base_url, timeout=timeout, max_retries=max_retries, logging=logging)
+ self.httpx_client = HttpClient(
+ httpx_client=httpx_client,
+ base_headers=self.get_headers,
+ base_timeout=self.get_timeout,
+ base_url=self.get_base_url,
+ base_max_retries=self.get_max_retries(),
+ logging_config=self._logging,
+ )
+
+
+class AsyncClientWrapper(BaseClientWrapper):
+ def __init__(
+ self,
+ *,
+ headers: typing.Optional[typing.Dict[str, str]] = None,
+ base_url: str,
+ timeout: typing.Optional[float] = None,
+ max_retries: int = 2,
+ logging: typing.Optional[typing.Union[LogConfig, Logger]] = None,
+ async_token: typing.Optional[typing.Callable[[], typing.Awaitable[str]]] = None,
+ httpx_client: httpx.AsyncClient,
+ ):
+ super().__init__(headers=headers, base_url=base_url, timeout=timeout, max_retries=max_retries, logging=logging)
+ self._async_token = async_token
+ self.httpx_client = AsyncHttpClient(
+ httpx_client=httpx_client,
+ base_headers=self.get_headers,
+ base_timeout=self.get_timeout,
+ base_url=self.get_base_url,
+ base_max_retries=self.get_max_retries(),
+ async_base_headers=self.async_get_headers,
+ logging_config=self._logging,
+ )
+
+ async def async_get_headers(self) -> typing.Dict[str, str]:
+ headers = self.get_headers()
+ if self._async_token is not None:
+ token = await self._async_token()
+ headers["Authorization"] = f"Bearer {token}"
+ return headers
diff --git a/sdks/python/core/datetime_utils.py b/sdks/python/core/datetime_utils.py
new file mode 100644
index 0000000000..a12b2ad03c
--- /dev/null
+++ b/sdks/python/core/datetime_utils.py
@@ -0,0 +1,70 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import datetime as dt
+from email.utils import parsedate_to_datetime
+from typing import Any
+
+import pydantic
+
+IS_PYDANTIC_V2 = pydantic.VERSION.startswith("2.")
+
+
+def parse_rfc2822_datetime(v: Any) -> dt.datetime:
+ """
+ Parse an RFC 2822 datetime string (e.g., "Wed, 02 Oct 2002 13:00:00 GMT")
+ into a datetime object. If the value is already a datetime, return it as-is.
+ Falls back to ISO 8601 parsing if RFC 2822 parsing fails.
+ """
+ if isinstance(v, dt.datetime):
+ return v
+ if isinstance(v, str):
+ try:
+ return parsedate_to_datetime(v)
+ except Exception:
+ pass
+ # Fallback to ISO 8601 parsing
+ return dt.datetime.fromisoformat(v.replace("Z", "+00:00"))
+ raise ValueError(f"Expected str or datetime, got {type(v)}")
+
+
+class Rfc2822DateTime(dt.datetime):
+ """A datetime subclass that parses RFC 2822 date strings.
+
+ On Pydantic V1, uses __get_validators__ for pre-validation.
+ On Pydantic V2, uses __get_pydantic_core_schema__ for BeforeValidator-style parsing.
+ """
+
+ @classmethod
+ def __get_validators__(cls): # type: ignore[no-untyped-def]
+ yield parse_rfc2822_datetime
+
+ @classmethod
+ def __get_pydantic_core_schema__(cls, _source_type: Any, _handler: Any) -> Any: # type: ignore[override]
+ from pydantic_core import core_schema
+
+ return core_schema.no_info_before_validator_function(parse_rfc2822_datetime, core_schema.datetime_schema())
+
+
+def serialize_datetime(v: dt.datetime) -> str:
+ """
+ Serialize a datetime including timezone info.
+
+ Uses the timezone info provided if present, otherwise uses the current runtime's timezone info.
+
+ UTC datetimes end in "Z" while all other timezones are represented as offset from UTC, e.g. +05:00.
+ """
+
+ def _serialize_zoned_datetime(v: dt.datetime) -> str:
+ if v.tzinfo is not None and v.tzinfo.tzname(None) == dt.timezone.utc.tzname(None):
+ # UTC is a special case where we use "Z" at the end instead of "+00:00"
+ return v.isoformat().replace("+00:00", "Z")
+ else:
+ # Delegate to the typical +/- offset format
+ return v.isoformat()
+
+ if v.tzinfo is not None:
+ return _serialize_zoned_datetime(v)
+ else:
+ local_tz = dt.datetime.now().astimezone().tzinfo
+ localized_dt = v.replace(tzinfo=local_tz)
+ return _serialize_zoned_datetime(localized_dt)
diff --git a/sdks/python/core/file.py b/sdks/python/core/file.py
new file mode 100644
index 0000000000..44b0d27c08
--- /dev/null
+++ b/sdks/python/core/file.py
@@ -0,0 +1,67 @@
+# This file was auto-generated by Fern from our API Definition.
+
+from typing import IO, Dict, List, Mapping, Optional, Tuple, Union, cast
+
+# File typing inspired by the flexibility of types within the httpx library
+# https://github.com/encode/httpx/blob/master/httpx/_types.py
+FileContent = Union[IO[bytes], bytes, str]
+File = Union[
+ # file (or bytes)
+ FileContent,
+ # (filename, file (or bytes))
+ Tuple[Optional[str], FileContent],
+ # (filename, file (or bytes), content_type)
+ Tuple[Optional[str], FileContent, Optional[str]],
+ # (filename, file (or bytes), content_type, headers)
+ Tuple[
+ Optional[str],
+ FileContent,
+ Optional[str],
+ Mapping[str, str],
+ ],
+]
+
+
+def convert_file_dict_to_httpx_tuples(
+ d: Dict[str, Union[File, List[File]]],
+) -> List[Tuple[str, File]]:
+ """
+ The format we use is a list of tuples, where the first element is the
+ name of the file and the second is the file object. Typically HTTPX wants
+ a dict, but to be able to send lists of files, you have to use the list
+ approach (which also works for non-lists)
+ https://github.com/encode/httpx/pull/1032
+ """
+
+ httpx_tuples = []
+ for key, file_like in d.items():
+ if isinstance(file_like, list):
+ for file_like_item in file_like:
+ httpx_tuples.append((key, file_like_item))
+ else:
+ httpx_tuples.append((key, file_like))
+ return httpx_tuples
+
+
+def with_content_type(*, file: File, default_content_type: str) -> File:
+ """
+ This function resolves to the file's content type, if provided, and defaults
+ to the default_content_type value if not.
+ """
+ if isinstance(file, tuple):
+ if len(file) == 2:
+ filename, content = cast(Tuple[Optional[str], FileContent], file) # type: ignore
+ return (filename, content, default_content_type)
+ elif len(file) == 3:
+ filename, content, file_content_type = cast(Tuple[Optional[str], FileContent, Optional[str]], file) # type: ignore
+ out_content_type = file_content_type or default_content_type
+ return (filename, content, out_content_type)
+ elif len(file) == 4:
+ filename, content, file_content_type, headers = cast( # type: ignore
+ Tuple[Optional[str], FileContent, Optional[str], Mapping[str, str]], file
+ )
+ out_content_type = file_content_type or default_content_type
+ return (filename, content, out_content_type, headers)
+ else:
+ raise ValueError(f"Unexpected tuple length: {len(file)}")
+ return (None, file, default_content_type)
diff --git a/sdks/python/core/force_multipart.py b/sdks/python/core/force_multipart.py
new file mode 100644
index 0000000000..5440913fd4
--- /dev/null
+++ b/sdks/python/core/force_multipart.py
@@ -0,0 +1,18 @@
+# This file was auto-generated by Fern from our API Definition.
+
+from typing import Any, Dict
+
+
+class ForceMultipartDict(Dict[str, Any]):
+ """
+ A dictionary subclass that always evaluates to True in boolean contexts.
+
+ This is used to force multipart/form-data encoding in HTTP requests even when
+ the dictionary is empty, which would normally evaluate to False.
+ """
+
+ def __bool__(self) -> bool:
+ return True
+
+
+FORCE_MULTIPART = ForceMultipartDict()
diff --git a/sdks/python/core/http_client.py b/sdks/python/core/http_client.py
new file mode 100644
index 0000000000..f686c5710d
--- /dev/null
+++ b/sdks/python/core/http_client.py
@@ -0,0 +1,839 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import asyncio
+import email.utils
+import re
+import time
+import typing
+from contextlib import asynccontextmanager, contextmanager
+from random import random
+
+import httpx
+from .file import File, convert_file_dict_to_httpx_tuples
+from .force_multipart import FORCE_MULTIPART
+from .jsonable_encoder import jsonable_encoder
+from .logging import LogConfig, Logger, create_logger
+from .query_encoder import encode_query
+from .remove_none_from_dict import remove_none_from_dict as remove_none_from_dict
+from .request_options import RequestOptions
+from httpx._types import RequestFiles
+
+INITIAL_RETRY_DELAY_SECONDS = 1.0
+MAX_RETRY_DELAY_SECONDS = 60.0
+JITTER_FACTOR = 0.2 # 20% random jitter
+
+
+def _parse_retry_after(response_headers: httpx.Headers) -> typing.Optional[float]:
+ """
+ This function parses the `Retry-After` header in a HTTP response and returns the number of seconds to wait.
+
+ Inspired by the urllib3 retry implementation.
+ """
+ retry_after_ms = response_headers.get("retry-after-ms")
+ if retry_after_ms is not None:
+ try:
+ return int(retry_after_ms) / 1000 if retry_after_ms > 0 else 0
+ except Exception:
+ pass
+
+ retry_after = response_headers.get("retry-after")
+ if retry_after is None:
+ return None
+
+ # Attempt to parse the header as an int.
+ if re.match(r"^\s*[0-9]+\s*$", retry_after):
+ seconds = float(retry_after)
+ # Fallback to parsing it as a date.
+ else:
+ retry_date_tuple = email.utils.parsedate_tz(retry_after)
+ if retry_date_tuple is None:
+ return None
+ if retry_date_tuple[9] is None: # Python 2
+ # Assume UTC if no timezone was specified
+ # On Python2.7, parsedate_tz returns None for a timezone offset
+ # instead of 0 if no timezone is given, where mktime_tz treats
+ # a None timezone offset as local time.
+ retry_date_tuple = retry_date_tuple[:9] + (0,) + retry_date_tuple[10:]
+
+ retry_date = email.utils.mktime_tz(retry_date_tuple)
+ seconds = retry_date - time.time()
+
+ if seconds < 0:
+ seconds = 0
+
+ return seconds
+
+
+def _add_positive_jitter(delay: float) -> float:
+ """Add positive jitter (0-20%) to prevent thundering herd."""
+ jitter_multiplier = 1 + random() * JITTER_FACTOR
+ return delay * jitter_multiplier
+
+
+def _add_symmetric_jitter(delay: float) -> float:
+ """Add symmetric jitter (±10%) for exponential backoff."""
+ jitter_multiplier = 1 + (random() - 0.5) * JITTER_FACTOR
+ return delay * jitter_multiplier
+
+
+def _parse_x_ratelimit_reset(response_headers: httpx.Headers) -> typing.Optional[float]:
+ """
+ Parse the X-RateLimit-Reset header (Unix timestamp in seconds).
+ Returns seconds to wait, or None if header is missing/invalid.
+ """
+ reset_time_str = response_headers.get("x-ratelimit-reset")
+ if reset_time_str is None:
+ return None
+
+ try:
+ reset_time = int(reset_time_str)
+ delay = reset_time - time.time()
+ if delay > 0:
+ return delay
+ except (ValueError, TypeError):
+ pass
+
+ return None
+
+
+def _retry_timeout(response: httpx.Response, retries: int) -> float:
+ """
+ Determine the amount of time to wait before retrying a request.
+ This function begins by trying to parse a retry-after header from the response, and then proceeds to use exponential backoff
+ with a jitter to determine the number of seconds to wait.
+ """
+
+ # 1. Check Retry-After header first
+ retry_after = _parse_retry_after(response.headers)
+ if retry_after is not None and retry_after > 0:
+ return min(retry_after, MAX_RETRY_DELAY_SECONDS)
+
+ # 2. Check X-RateLimit-Reset header (with positive jitter)
+ ratelimit_reset = _parse_x_ratelimit_reset(response.headers)
+ if ratelimit_reset is not None:
+ return _add_positive_jitter(min(ratelimit_reset, MAX_RETRY_DELAY_SECONDS))
+
+ # 3. Fall back to exponential backoff (with symmetric jitter)
+ backoff = min(INITIAL_RETRY_DELAY_SECONDS * pow(2.0, retries), MAX_RETRY_DELAY_SECONDS)
+ return _add_symmetric_jitter(backoff)
+
+
+def _retry_timeout_from_retries(retries: int) -> float:
+ """Determine retry timeout using exponential backoff when no response is available."""
+ backoff = min(INITIAL_RETRY_DELAY_SECONDS * pow(2.0, retries), MAX_RETRY_DELAY_SECONDS)
+ return _add_symmetric_jitter(backoff)
+
+
+def _should_retry(response: httpx.Response) -> bool:
+ return response.status_code >= 500 or response.status_code in [429, 408, 409]
+
+
+_SENSITIVE_HEADERS = frozenset(
+ {
+ "authorization",
+ "www-authenticate",
+ "x-api-key",
+ "api-key",
+ "apikey",
+ "x-api-token",
+ "x-auth-token",
+ "auth-token",
+ "cookie",
+ "set-cookie",
+ "proxy-authorization",
+ "proxy-authenticate",
+ "x-csrf-token",
+ "x-xsrf-token",
+ "x-session-token",
+ "x-access-token",
+ }
+)
+
+
+def _redact_headers(headers: typing.Dict[str, str]) -> typing.Dict[str, str]:
+ return {k: ("[REDACTED]" if k.lower() in _SENSITIVE_HEADERS else v) for k, v in headers.items()}
+
+
+def _build_url(base_url: str, path: typing.Optional[str]) -> str:
+ """
+ Build a full URL by joining a base URL with a path.
+
+ This function correctly handles base URLs that contain path prefixes (e.g., tenant-based URLs)
+ by using string concatenation instead of urllib.parse.urljoin(), which would incorrectly
+ strip path components when the path starts with '/'.
+
+ Example:
+ >>> _build_url("https://cloud.example.com/org/tenant/api", "/users")
+ 'https://cloud.example.com/org/tenant/api/users'
+
+ Args:
+ base_url: The base URL, which may contain path prefixes.
+ path: The path to append. Can be None or empty string.
+
+ Returns:
+ The full URL with base_url and path properly joined.
+ """
+ if not path:
+ return base_url
+ return f"{base_url.rstrip('/')}/{path.lstrip('/')}"
+
+
+def _maybe_filter_none_from_multipart_data(
+ data: typing.Optional[typing.Any],
+ request_files: typing.Optional[RequestFiles],
+ force_multipart: typing.Optional[bool],
+) -> typing.Optional[typing.Any]:
+ """
+ Filter None values from data body for multipart/form requests.
+ This prevents httpx from converting None to empty strings in multipart encoding.
+ Only applies when files are present or force_multipart is True.
+ """
+ if data is not None and isinstance(data, typing.Mapping) and (request_files or force_multipart):
+ return remove_none_from_dict(data)
+ return data
+
+
+def remove_omit_from_dict(
+ original: typing.Dict[str, typing.Optional[typing.Any]],
+ omit: typing.Optional[typing.Any],
+) -> typing.Dict[str, typing.Any]:
+ if omit is None:
+ return original
+ new: typing.Dict[str, typing.Any] = {}
+ for key, value in original.items():
+ if value is not omit:
+ new[key] = value
+ return new
+
+
+def maybe_filter_request_body(
+ data: typing.Optional[typing.Any],
+ request_options: typing.Optional[RequestOptions],
+ omit: typing.Optional[typing.Any],
+) -> typing.Optional[typing.Any]:
+ if data is None:
+ return (
+ jsonable_encoder(request_options.get("additional_body_parameters", {})) or {}
+ if request_options is not None
+ else None
+ )
+ elif not isinstance(data, typing.Mapping):
+ data_content = jsonable_encoder(data)
+ else:
+ data_content = {
+ **(jsonable_encoder(remove_omit_from_dict(data, omit))), # type: ignore
+ **(
+ jsonable_encoder(request_options.get("additional_body_parameters", {})) or {}
+ if request_options is not None
+ else {}
+ ),
+ }
+ return data_content
+
+
+# Abstracted out for testing purposes
+def get_request_body(
+ *,
+ json: typing.Optional[typing.Any],
+ data: typing.Optional[typing.Any],
+ request_options: typing.Optional[RequestOptions],
+ omit: typing.Optional[typing.Any],
+) -> typing.Tuple[typing.Optional[typing.Any], typing.Optional[typing.Any]]:
+ json_body = None
+ data_body = None
+ if data is not None:
+ data_body = maybe_filter_request_body(data, request_options, omit)
+ else:
+ # If both data and json are None, we send json data in the event extra properties are specified
+ json_body = maybe_filter_request_body(json, request_options, omit)
+
+ has_additional_body_parameters = bool(
+ request_options is not None and request_options.get("additional_body_parameters")
+ )
+
+ # Only collapse empty dict to None when the body was not explicitly provided
+ # and there are no additional body parameters. This preserves explicit empty
+ # bodies (e.g., when an endpoint has a request body type but all fields are optional).
+ if json_body == {} and json is None and not has_additional_body_parameters:
+ json_body = None
+ if data_body == {} and data is None and not has_additional_body_parameters:
+ data_body = None
+
+ return json_body, data_body
+
+
+class HttpClient:
+ def __init__(
+ self,
+ *,
+ httpx_client: httpx.Client,
+ base_timeout: typing.Callable[[], typing.Optional[float]],
+ base_headers: typing.Callable[[], typing.Dict[str, str]],
+ base_url: typing.Optional[typing.Callable[[], str]] = None,
+ base_max_retries: int = 2,
+ logging_config: typing.Optional[typing.Union[LogConfig, Logger]] = None,
+ ):
+ self.base_url = base_url
+ self.base_timeout = base_timeout
+ self.base_headers = base_headers
+ self.base_max_retries = base_max_retries
+ self.httpx_client = httpx_client
+ self.logger = create_logger(logging_config)
+
+ def get_base_url(self, maybe_base_url: typing.Optional[str]) -> str:
+ base_url = maybe_base_url
+ if self.base_url is not None and base_url is None:
+ base_url = self.base_url()
+
+ if base_url is None:
+ raise ValueError("A base_url is required to make this request, please provide one and try again.")
+ return base_url
+
+ def request(
+ self,
+ path: typing.Optional[str] = None,
+ *,
+ method: str,
+ base_url: typing.Optional[str] = None,
+ params: typing.Optional[typing.Dict[str, typing.Any]] = None,
+ json: typing.Optional[typing.Any] = None,
+ data: typing.Optional[typing.Any] = None,
+ content: typing.Optional[typing.Union[bytes, typing.Iterator[bytes], typing.AsyncIterator[bytes]]] = None,
+ files: typing.Optional[
+ typing.Union[
+ typing.Dict[str, typing.Optional[typing.Union[File, typing.List[File]]]],
+ typing.List[typing.Tuple[str, File]],
+ ]
+ ] = None,
+ headers: typing.Optional[typing.Dict[str, typing.Any]] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ retries: int = 0,
+ omit: typing.Optional[typing.Any] = None,
+ force_multipart: typing.Optional[bool] = None,
+ ) -> httpx.Response:
+ base_url = self.get_base_url(base_url)
+ timeout = (
+ request_options.get("timeout_in_seconds")
+ if request_options is not None and request_options.get("timeout_in_seconds") is not None
+ else self.base_timeout()
+ )
+
+ json_body, data_body = get_request_body(json=json, data=data, request_options=request_options, omit=omit)
+
+ request_files: typing.Optional[RequestFiles] = (
+ convert_file_dict_to_httpx_tuples(remove_omit_from_dict(remove_none_from_dict(files), omit))
+ if (files is not None and files is not omit and isinstance(files, dict))
+ else None
+ )
+
+ if (request_files is None or len(request_files) == 0) and force_multipart:
+ request_files = FORCE_MULTIPART
+
+ data_body = _maybe_filter_none_from_multipart_data(data_body, request_files, force_multipart)
+
+ # Compute encoded params separately to avoid passing empty list to httpx
+ # (httpx strips existing query params from URL when params=[] is passed)
+ _encoded_params = encode_query(
+ jsonable_encoder(
+ remove_none_from_dict(
+ remove_omit_from_dict(
+ {
+ **(params if params is not None else {}),
+ **(
+ request_options.get("additional_query_parameters", {}) or {}
+ if request_options is not None
+ else {}
+ ),
+ },
+ omit,
+ )
+ )
+ )
+ )
+
+ _request_url = _build_url(base_url, path)
+ _request_headers = jsonable_encoder(
+ remove_none_from_dict(
+ {
+ **self.base_headers(),
+ **(headers if headers is not None else {}),
+ **(request_options.get("additional_headers", {}) or {} if request_options is not None else {}),
+ }
+ )
+ )
+
+ if self.logger.is_debug():
+ self.logger.debug(
+ "Making HTTP request",
+ method=method,
+ url=_request_url,
+ headers=_redact_headers(_request_headers),
+ has_body=json_body is not None or data_body is not None,
+ )
+
+ max_retries: int = (
+ request_options.get("max_retries", self.base_max_retries)
+ if request_options is not None
+ else self.base_max_retries
+ )
+
+ try:
+ response = self.httpx_client.request(
+ method=method,
+ url=_request_url,
+ headers=_request_headers,
+ params=_encoded_params if _encoded_params else None,
+ json=json_body,
+ data=data_body,
+ content=content,
+ files=request_files,
+ timeout=timeout,
+ )
+ except (httpx.ConnectError, httpx.RemoteProtocolError):
+ if retries < max_retries:
+ time.sleep(_retry_timeout_from_retries(retries=retries))
+ return self.request(
+ path=path,
+ method=method,
+ base_url=base_url,
+ params=params,
+ json=json,
+ data=data,
+ content=content,
+ files=files,
+ headers=headers,
+ request_options=request_options,
+ retries=retries + 1,
+ omit=omit,
+ force_multipart=force_multipart,
+ )
+ raise
+
+ if _should_retry(response=response):
+ if retries < max_retries:
+ time.sleep(_retry_timeout(response=response, retries=retries))
+ return self.request(
+ path=path,
+ method=method,
+ base_url=base_url,
+ params=params,
+ json=json,
+ data=data,
+ content=content,
+ files=files,
+ headers=headers,
+ request_options=request_options,
+ retries=retries + 1,
+ omit=omit,
+ force_multipart=force_multipart,
+ )
+
+ if self.logger.is_debug():
+ if 200 <= response.status_code < 400:
+ self.logger.debug(
+ "HTTP request succeeded",
+ method=method,
+ url=_request_url,
+ status_code=response.status_code,
+ )
+
+ if self.logger.is_error():
+ if response.status_code >= 400:
+ self.logger.error(
+ "HTTP request failed with error status",
+ method=method,
+ url=_request_url,
+ status_code=response.status_code,
+ )
+
+ return response
+
+ @contextmanager
+ def stream(
+ self,
+ path: typing.Optional[str] = None,
+ *,
+ method: str,
+ base_url: typing.Optional[str] = None,
+ params: typing.Optional[typing.Dict[str, typing.Any]] = None,
+ json: typing.Optional[typing.Any] = None,
+ data: typing.Optional[typing.Any] = None,
+ content: typing.Optional[typing.Union[bytes, typing.Iterator[bytes], typing.AsyncIterator[bytes]]] = None,
+ files: typing.Optional[
+ typing.Union[
+ typing.Dict[str, typing.Optional[typing.Union[File, typing.List[File]]]],
+ typing.List[typing.Tuple[str, File]],
+ ]
+ ] = None,
+ headers: typing.Optional[typing.Dict[str, typing.Any]] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ retries: int = 0,
+ omit: typing.Optional[typing.Any] = None,
+ force_multipart: typing.Optional[bool] = None,
+ ) -> typing.Iterator[httpx.Response]:
+ base_url = self.get_base_url(base_url)
+ timeout = (
+ request_options.get("timeout_in_seconds")
+ if request_options is not None and request_options.get("timeout_in_seconds") is not None
+ else self.base_timeout()
+ )
+
+ request_files: typing.Optional[RequestFiles] = (
+ convert_file_dict_to_httpx_tuples(remove_omit_from_dict(remove_none_from_dict(files), omit))
+ if (files is not None and files is not omit and isinstance(files, dict))
+ else None
+ )
+
+ if (request_files is None or len(request_files) == 0) and force_multipart:
+ request_files = FORCE_MULTIPART
+
+ json_body, data_body = get_request_body(json=json, data=data, request_options=request_options, omit=omit)
+
+ data_body = _maybe_filter_none_from_multipart_data(data_body, request_files, force_multipart)
+
+ # Compute encoded params separately to avoid passing empty list to httpx
+ # (httpx strips existing query params from URL when params=[] is passed)
+ _encoded_params = encode_query(
+ jsonable_encoder(
+ remove_none_from_dict(
+ remove_omit_from_dict(
+ {
+ **(params if params is not None else {}),
+ **(
+ request_options.get("additional_query_parameters", {})
+ if request_options is not None
+ else {}
+ ),
+ },
+ omit,
+ )
+ )
+ )
+ )
+
+ _request_url = _build_url(base_url, path)
+ _request_headers = jsonable_encoder(
+ remove_none_from_dict(
+ {
+ **self.base_headers(),
+ **(headers if headers is not None else {}),
+ **(request_options.get("additional_headers", {}) if request_options is not None else {}),
+ }
+ )
+ )
+
+ if self.logger.is_debug():
+ self.logger.debug(
+ "Making streaming HTTP request",
+ method=method,
+ url=_request_url,
+ headers=_redact_headers(_request_headers),
+ )
+
+ with self.httpx_client.stream(
+ method=method,
+ url=_request_url,
+ headers=_request_headers,
+ params=_encoded_params if _encoded_params else None,
+ json=json_body,
+ data=data_body,
+ content=content,
+ files=request_files,
+ timeout=timeout,
+ ) as stream:
+ yield stream
+
+
+class AsyncHttpClient:
+ def __init__(
+ self,
+ *,
+ httpx_client: httpx.AsyncClient,
+ base_timeout: typing.Callable[[], typing.Optional[float]],
+ base_headers: typing.Callable[[], typing.Dict[str, str]],
+ base_url: typing.Optional[typing.Callable[[], str]] = None,
+ base_max_retries: int = 2,
+ async_base_headers: typing.Optional[typing.Callable[[], typing.Awaitable[typing.Dict[str, str]]]] = None,
+ logging_config: typing.Optional[typing.Union[LogConfig, Logger]] = None,
+ ):
+ self.base_url = base_url
+ self.base_timeout = base_timeout
+ self.base_headers = base_headers
+ self.base_max_retries = base_max_retries
+ self.async_base_headers = async_base_headers
+ self.httpx_client = httpx_client
+ self.logger = create_logger(logging_config)
+
+ async def _get_headers(self) -> typing.Dict[str, str]:
+ if self.async_base_headers is not None:
+ return await self.async_base_headers()
+ return self.base_headers()
+
+ def get_base_url(self, maybe_base_url: typing.Optional[str]) -> str:
+ base_url = maybe_base_url
+ if self.base_url is not None and base_url is None:
+ base_url = self.base_url()
+
+ if base_url is None:
+ raise ValueError("A base_url is required to make this request, please provide one and try again.")
+ return base_url
+
+ async def request(
+ self,
+ path: typing.Optional[str] = None,
+ *,
+ method: str,
+ base_url: typing.Optional[str] = None,
+ params: typing.Optional[typing.Dict[str, typing.Any]] = None,
+ json: typing.Optional[typing.Any] = None,
+ data: typing.Optional[typing.Any] = None,
+ content: typing.Optional[typing.Union[bytes, typing.Iterator[bytes], typing.AsyncIterator[bytes]]] = None,
+ files: typing.Optional[
+ typing.Union[
+ typing.Dict[str, typing.Optional[typing.Union[File, typing.List[File]]]],
+ typing.List[typing.Tuple[str, File]],
+ ]
+ ] = None,
+ headers: typing.Optional[typing.Dict[str, typing.Any]] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ retries: int = 0,
+ omit: typing.Optional[typing.Any] = None,
+ force_multipart: typing.Optional[bool] = None,
+ ) -> httpx.Response:
+ base_url = self.get_base_url(base_url)
+ timeout = (
+ request_options.get("timeout_in_seconds")
+ if request_options is not None and request_options.get("timeout_in_seconds") is not None
+ else self.base_timeout()
+ )
+
+ request_files: typing.Optional[RequestFiles] = (
+ convert_file_dict_to_httpx_tuples(remove_omit_from_dict(remove_none_from_dict(files), omit))
+ if (files is not None and files is not omit and isinstance(files, dict))
+ else None
+ )
+
+ if (request_files is None or len(request_files) == 0) and force_multipart:
+ request_files = FORCE_MULTIPART
+
+ json_body, data_body = get_request_body(json=json, data=data, request_options=request_options, omit=omit)
+
+ data_body = _maybe_filter_none_from_multipart_data(data_body, request_files, force_multipart)
+
+ # Get headers (supports async token providers)
+ _headers = await self._get_headers()
+
+ # Compute encoded params separately to avoid passing empty list to httpx
+ # (httpx strips existing query params from URL when params=[] is passed)
+ _encoded_params = encode_query(
+ jsonable_encoder(
+ remove_none_from_dict(
+ remove_omit_from_dict(
+ {
+ **(params if params is not None else {}),
+ **(
+ request_options.get("additional_query_parameters", {}) or {}
+ if request_options is not None
+ else {}
+ ),
+ },
+ omit,
+ )
+ )
+ )
+ )
+
+ _request_url = _build_url(base_url, path)
+ _request_headers = jsonable_encoder(
+ remove_none_from_dict(
+ {
+ **_headers,
+ **(headers if headers is not None else {}),
+ **(request_options.get("additional_headers", {}) or {} if request_options is not None else {}),
+ }
+ )
+ )
+
+ if self.logger.is_debug():
+ self.logger.debug(
+ "Making HTTP request",
+ method=method,
+ url=_request_url,
+ headers=_redact_headers(_request_headers),
+ has_body=json_body is not None or data_body is not None,
+ )
+
+ max_retries: int = (
+ request_options.get("max_retries", self.base_max_retries)
+ if request_options is not None
+ else self.base_max_retries
+ )
+
+ try:
+ response = await self.httpx_client.request(
+ method=method,
+ url=_request_url,
+ headers=_request_headers,
+ params=_encoded_params if _encoded_params else None,
+ json=json_body,
+ data=data_body,
+ content=content,
+ files=request_files,
+ timeout=timeout,
+ )
+ except (httpx.ConnectError, httpx.RemoteProtocolError):
+ if retries < max_retries:
+ await asyncio.sleep(_retry_timeout_from_retries(retries=retries))
+ return await self.request(
+ path=path,
+ method=method,
+ base_url=base_url,
+ params=params,
+ json=json,
+ data=data,
+ content=content,
+ files=files,
+ headers=headers,
+ request_options=request_options,
+ retries=retries + 1,
+ omit=omit,
+ force_multipart=force_multipart,
+ )
+ raise
+
+ if _should_retry(response=response):
+ if retries < max_retries:
+ await asyncio.sleep(_retry_timeout(response=response, retries=retries))
+ return await self.request(
+ path=path,
+ method=method,
+ base_url=base_url,
+ params=params,
+ json=json,
+ data=data,
+ content=content,
+ files=files,
+ headers=headers,
+ request_options=request_options,
+ retries=retries + 1,
+ omit=omit,
+ force_multipart=force_multipart,
+ )
+
+ if self.logger.is_debug():
+ if 200 <= response.status_code < 400:
+ self.logger.debug(
+ "HTTP request succeeded",
+ method=method,
+ url=_request_url,
+ status_code=response.status_code,
+ )
+
+ if self.logger.is_error():
+ if response.status_code >= 400:
+ self.logger.error(
+ "HTTP request failed with error status",
+ method=method,
+ url=_request_url,
+ status_code=response.status_code,
+ )
+
+ return response
+
+ @asynccontextmanager
+ async def stream(
+ self,
+ path: typing.Optional[str] = None,
+ *,
+ method: str,
+ base_url: typing.Optional[str] = None,
+ params: typing.Optional[typing.Dict[str, typing.Any]] = None,
+ json: typing.Optional[typing.Any] = None,
+ data: typing.Optional[typing.Any] = None,
+ content: typing.Optional[typing.Union[bytes, typing.Iterator[bytes], typing.AsyncIterator[bytes]]] = None,
+ files: typing.Optional[
+ typing.Union[
+ typing.Dict[str, typing.Optional[typing.Union[File, typing.List[File]]]],
+ typing.List[typing.Tuple[str, File]],
+ ]
+ ] = None,
+ headers: typing.Optional[typing.Dict[str, typing.Any]] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ retries: int = 0,
+ omit: typing.Optional[typing.Any] = None,
+ force_multipart: typing.Optional[bool] = None,
+ ) -> typing.AsyncIterator[httpx.Response]:
+ base_url = self.get_base_url(base_url)
+ timeout = (
+ request_options.get("timeout_in_seconds")
+ if request_options is not None and request_options.get("timeout_in_seconds") is not None
+ else self.base_timeout()
+ )
+
+ request_files: typing.Optional[RequestFiles] = (
+ convert_file_dict_to_httpx_tuples(remove_omit_from_dict(remove_none_from_dict(files), omit))
+ if (files is not None and files is not omit and isinstance(files, dict))
+ else None
+ )
+
+ if (request_files is None or len(request_files) == 0) and force_multipart:
+ request_files = FORCE_MULTIPART
+
+ json_body, data_body = get_request_body(json=json, data=data, request_options=request_options, omit=omit)
+
+ data_body = _maybe_filter_none_from_multipart_data(data_body, request_files, force_multipart)
+
+ # Get headers (supports async token providers)
+ _headers = await self._get_headers()
+
+ # Compute encoded params separately to avoid passing empty list to httpx
+ # (httpx strips existing query params from URL when params=[] is passed)
+ _encoded_params = encode_query(
+ jsonable_encoder(
+ remove_none_from_dict(
+ remove_omit_from_dict(
+ {
+ **(params if params is not None else {}),
+ **(
+ request_options.get("additional_query_parameters", {})
+ if request_options is not None
+ else {}
+ ),
+ },
+ omit=omit,
+ )
+ )
+ )
+ )
+
+ _request_url = _build_url(base_url, path)
+ _request_headers = jsonable_encoder(
+ remove_none_from_dict(
+ {
+ **_headers,
+ **(headers if headers is not None else {}),
+ **(request_options.get("additional_headers", {}) if request_options is not None else {}),
+ }
+ )
+ )
+
+ if self.logger.is_debug():
+ self.logger.debug(
+ "Making streaming HTTP request",
+ method=method,
+ url=_request_url,
+ headers=_redact_headers(_request_headers),
+ )
+
+ async with self.httpx_client.stream(
+ method=method,
+ url=_request_url,
+ headers=_request_headers,
+ params=_encoded_params if _encoded_params else None,
+ json=json_body,
+ data=data_body,
+ content=content,
+ files=request_files,
+ timeout=timeout,
+ ) as stream:
+ yield stream
diff --git a/sdks/python/core/http_response.py b/sdks/python/core/http_response.py
new file mode 100644
index 0000000000..00bb1096d2
--- /dev/null
+++ b/sdks/python/core/http_response.py
@@ -0,0 +1,59 @@
+# This file was auto-generated by Fern from our API Definition.
+
+from typing import Dict, Generic, TypeVar
+
+import httpx
+
+# Generic to represent the underlying type of the data wrapped by the HTTP response.
+T = TypeVar("T")
+
+
+class BaseHttpResponse:
+ """Minimalist HTTP response wrapper that exposes response headers and status code."""
+
+ _response: httpx.Response
+
+ def __init__(self, response: httpx.Response):
+ self._response = response
+
+ @property
+ def headers(self) -> Dict[str, str]:
+ return dict(self._response.headers)
+
+ @property
+ def status_code(self) -> int:
+ return self._response.status_code
+
+
+class HttpResponse(Generic[T], BaseHttpResponse):
+ """HTTP response wrapper that exposes response headers and data."""
+
+ _data: T
+
+ def __init__(self, response: httpx.Response, data: T):
+ super().__init__(response)
+ self._data = data
+
+ @property
+ def data(self) -> T:
+ return self._data
+
+ def close(self) -> None:
+ self._response.close()
+
+
+class AsyncHttpResponse(Generic[T], BaseHttpResponse):
+ """HTTP response wrapper that exposes response headers and data."""
+
+ _data: T
+
+ def __init__(self, response: httpx.Response, data: T):
+ super().__init__(response)
+ self._data = data
+
+ @property
+ def data(self) -> T:
+ return self._data
+
+ async def close(self) -> None:
+ await self._response.aclose()
diff --git a/sdks/python/core/http_sse/__init__.py b/sdks/python/core/http_sse/__init__.py
new file mode 100644
index 0000000000..730e5a3382
--- /dev/null
+++ b/sdks/python/core/http_sse/__init__.py
@@ -0,0 +1,42 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
+import typing
+from importlib import import_module
+
+if typing.TYPE_CHECKING:
+ from ._api import EventSource, aconnect_sse, connect_sse
+ from ._exceptions import SSEError
+ from ._models import ServerSentEvent
+_dynamic_imports: typing.Dict[str, str] = {
+ "EventSource": "._api",
+ "SSEError": "._exceptions",
+ "ServerSentEvent": "._models",
+ "aconnect_sse": "._api",
+ "connect_sse": "._api",
+}
+
+
+def __getattr__(attr_name: str) -> typing.Any:
+ module_name = _dynamic_imports.get(attr_name)
+ if module_name is None:
+ raise AttributeError(f"No {attr_name} found in _dynamic_imports for module name -> {__name__}")
+ try:
+ module = import_module(module_name, __package__)
+ if module_name == f".{attr_name}":
+ return module
+ else:
+ return getattr(module, attr_name)
+ except ImportError as e:
+ raise ImportError(f"Failed to import {attr_name} from {module_name}: {e}") from e
+ except AttributeError as e:
+ raise AttributeError(f"Failed to get {attr_name} from {module_name}: {e}") from e
+
+
+def __dir__():
+ lazy_attrs = list(_dynamic_imports.keys())
+ return sorted(lazy_attrs)
+
+
+__all__ = ["EventSource", "SSEError", "ServerSentEvent", "aconnect_sse", "connect_sse"]
diff --git a/sdks/python/core/http_sse/_api.py b/sdks/python/core/http_sse/_api.py
new file mode 100644
index 0000000000..fd137301b9
--- /dev/null
+++ b/sdks/python/core/http_sse/_api.py
@@ -0,0 +1,170 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import codecs
+import re
+from contextlib import asynccontextmanager, contextmanager
+from typing import Any, AsyncGenerator, AsyncIterator, Iterator
+
+import httpx
+from ._decoders import SSEDecoder
+from ._exceptions import SSEError
+from ._models import ServerSentEvent
+
+MAX_LINE_SIZE: int = 1_048_576 # 1 MiB
+
+
+class EventSource:
+ def __init__(self, response: httpx.Response) -> None:
+ self._response = response
+
+ def _check_content_type(self) -> None:
+ content_type = self._response.headers.get("content-type", "").partition(";")[0]
+ if "text/event-stream" not in content_type:
+ raise SSEError(
+ f"Expected response header Content-Type to contain 'text/event-stream', got {content_type!r}"
+ )
+
+ def _get_charset(self) -> str:
+ """Extract charset from Content-Type header, fallback to UTF-8."""
+ content_type = self._response.headers.get("content-type", "")
+
+ # Parse charset parameter using regex
+ charset_match = re.search(r"charset=([^;\s]+)", content_type, re.IGNORECASE)
+ if charset_match:
+ charset = charset_match.group(1).strip("\"'")
+ # Validate that it's a known encoding
+ try:
+ # Test if the charset is valid by trying to encode/decode
+ "test".encode(charset).decode(charset)
+ return charset
+ except (LookupError, UnicodeError):
+ # If charset is invalid, fall back to UTF-8
+ pass
+
+ # Default to UTF-8 if no charset specified or invalid charset
+ return "utf-8"
+
+ @property
+ def response(self) -> httpx.Response:
+ return self._response
+
+ @staticmethod
+ def _normalize_sse_line_endings(buf: str) -> str:
+ """Normalize line endings per the SSE spec (\\r\\n → \\n, bare \\r → \\n).
+
+ A trailing \\r is preserved because it may pair with a leading \\n in
+ the next chunk to form a single \\r\\n terminator.
+ """
+ buf = buf.replace("\r\n", "\n")
+ if buf.endswith("\r"):
+ return buf[:-1].replace("\r", "\n") + "\r"
+ return buf.replace("\r", "\n")
+
+ def iter_sse(self) -> Iterator[ServerSentEvent]:
+ self._check_content_type()
+ decoder = SSEDecoder()
+ charset = self._get_charset()
+ text_decoder = codecs.getincrementaldecoder(charset)(errors="replace")
+
+ buf = ""
+ for chunk in self._response.iter_bytes():
+ buf += text_decoder.decode(chunk)
+ buf = self._normalize_sse_line_endings(buf)
+
+ while "\n" in buf:
+ line, buf = buf.split("\n", 1)
+ sse = decoder.decode(line)
+ if sse is not None:
+ yield sse
+
+ if len(buf) > MAX_LINE_SIZE:
+ raise SSEError(
+ f"SSE line exceeded maximum size of {MAX_LINE_SIZE} characters without encountering a newline"
+ )
+
+ # Flush any remaining bytes from the incremental decoder
+ buf += text_decoder.decode(b"", final=True)
+ buf = buf.replace("\r\n", "\n").replace("\r", "\n")
+
+ if len(buf) > MAX_LINE_SIZE:
+ raise SSEError(
+ f"SSE line exceeded maximum size of {MAX_LINE_SIZE} characters without encountering a newline"
+ )
+
+ while "\n" in buf:
+ line, buf = buf.split("\n", 1)
+ sse = decoder.decode(line)
+ if sse is not None:
+ yield sse
+
+ if buf.strip():
+ sse = decoder.decode(buf)
+ if sse is not None:
+ yield sse
+
+ async def aiter_sse(self) -> AsyncGenerator[ServerSentEvent, None]:
+ self._check_content_type()
+ decoder = SSEDecoder()
+ charset = self._get_charset()
+ text_decoder = codecs.getincrementaldecoder(charset)(errors="replace")
+
+ buf = ""
+ async for chunk in self._response.aiter_bytes():
+ buf += text_decoder.decode(chunk)
+ buf = self._normalize_sse_line_endings(buf)
+
+ while "\n" in buf:
+ line, buf = buf.split("\n", 1)
+ sse = decoder.decode(line)
+ if sse is not None:
+ yield sse
+
+ if len(buf) > MAX_LINE_SIZE:
+ raise SSEError(
+ f"SSE line exceeded maximum size of {MAX_LINE_SIZE} characters without encountering a newline"
+ )
+
+ # Flush any remaining bytes from the incremental decoder
+ buf += text_decoder.decode(b"", final=True)
+ buf = buf.replace("\r\n", "\n").replace("\r", "\n")
+
+ if len(buf) > MAX_LINE_SIZE:
+ raise SSEError(
+ f"SSE line exceeded maximum size of {MAX_LINE_SIZE} characters without encountering a newline"
+ )
+
+ while "\n" in buf:
+ line, buf = buf.split("\n", 1)
+ sse = decoder.decode(line)
+ if sse is not None:
+ yield sse
+
+ if buf.strip():
+ sse = decoder.decode(buf)
+ if sse is not None:
+ yield sse
+
+
+@contextmanager
+def connect_sse(client: httpx.Client, method: str, url: str, **kwargs: Any) -> Iterator[EventSource]:
+ headers = kwargs.pop("headers", {})
+ headers["Accept"] = "text/event-stream"
+ headers["Cache-Control"] = "no-store"
+
+ with client.stream(method, url, headers=headers, **kwargs) as response:
+ yield EventSource(response)
+
+
+@asynccontextmanager
+async def aconnect_sse(
+ client: httpx.AsyncClient,
+ method: str,
+ url: str,
+ **kwargs: Any,
+) -> AsyncIterator[EventSource]:
+ headers = kwargs.pop("headers", {})
+ headers["Accept"] = "text/event-stream"
+ headers["Cache-Control"] = "no-store"
+
+ async with client.stream(method, url, headers=headers, **kwargs) as response:
+ yield EventSource(response)
diff --git a/sdks/python/core/http_sse/_decoders.py b/sdks/python/core/http_sse/_decoders.py
new file mode 100644
index 0000000000..339b089013
--- /dev/null
+++ b/sdks/python/core/http_sse/_decoders.py
@@ -0,0 +1,61 @@
+# This file was auto-generated by Fern from our API Definition.
+
+from typing import List, Optional
+
+from ._models import ServerSentEvent
+
+
+class SSEDecoder:
+ def __init__(self) -> None:
+ self._event = ""
+ self._data: List[str] = []
+ self._last_event_id = ""
+ self._retry: Optional[int] = None
+
+ def decode(self, line: str) -> Optional[ServerSentEvent]:
+ # See: https://html.spec.whatwg.org/multipage/server-sent-events.html#event-stream-interpretation # noqa: E501
+
+ if not line:
+ if not self._event and not self._data and not self._last_event_id and self._retry is None:
+ return None
+
+ sse = ServerSentEvent(
+ event=self._event,
+ data="\n".join(self._data),
+ id=self._last_event_id,
+ retry=self._retry,
+ )
+
+ # NOTE: as per the SSE spec, do not reset last_event_id.
+ self._event = ""
+ self._data = []
+ self._retry = None
+
+ return sse
+
+ if line.startswith(":"):
+ return None
+
+ fieldname, _, value = line.partition(":")
+
+ if value.startswith(" "):
+ value = value[1:]
+
+ if fieldname == "event":
+ self._event = value
+ elif fieldname == "data":
+ self._data.append(value)
+ elif fieldname == "id":
+ if "\0" in value:
+ pass
+ else:
+ self._last_event_id = value
+ elif fieldname == "retry":
+ try:
+ self._retry = int(value)
+ except (TypeError, ValueError):
+ pass
+ else:
+ pass # Field is ignored.
+
+ return None
diff --git a/sdks/python/core/http_sse/_exceptions.py b/sdks/python/core/http_sse/_exceptions.py
new file mode 100644
index 0000000000..81605a8a65
--- /dev/null
+++ b/sdks/python/core/http_sse/_exceptions.py
@@ -0,0 +1,7 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import httpx
+
+
+class SSEError(httpx.TransportError):
+ pass
diff --git a/sdks/python/core/http_sse/_models.py b/sdks/python/core/http_sse/_models.py
new file mode 100644
index 0000000000..1af57f8fd0
--- /dev/null
+++ b/sdks/python/core/http_sse/_models.py
@@ -0,0 +1,17 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import json
+from dataclasses import dataclass
+from typing import Any, Optional
+
+
+@dataclass(frozen=True)
+class ServerSentEvent:
+ event: str = "message"
+ data: str = ""
+ id: str = ""
+ retry: Optional[int] = None
+
+ def json(self) -> Any:
+ """Parse the data field as JSON."""
+ return json.loads(self.data)
diff --git a/sdks/python/core/jsonable_encoder.py b/sdks/python/core/jsonable_encoder.py
new file mode 100644
index 0000000000..5b0902ebcd
--- /dev/null
+++ b/sdks/python/core/jsonable_encoder.py
@@ -0,0 +1,120 @@
+# This file was auto-generated by Fern from our API Definition.
+
+"""
+jsonable_encoder converts a Python object to a JSON-friendly dict
+(e.g. datetimes to strings, Pydantic models to dicts).
+
+Taken from FastAPI, and made a bit simpler
+https://github.com/tiangolo/fastapi/blob/master/fastapi/encoders.py
+"""
+
+import base64
+import dataclasses
+import datetime as dt
+from enum import Enum
+from pathlib import PurePath
+from types import GeneratorType
+from typing import Any, Callable, Dict, List, Optional, Set, Union
+
+import pydantic
+from .datetime_utils import serialize_datetime
+from .pydantic_utilities import (
+ IS_PYDANTIC_V2,
+ encode_by_type,
+ to_jsonable_with_fallback,
+)
+
+SetIntStr = Set[Union[int, str]]
+DictIntStrAny = Dict[Union[int, str], Any]
+
+
+def jsonable_encoder(obj: Any, custom_encoder: Optional[Dict[Any, Callable[[Any], Any]]] = None) -> Any:
+ custom_encoder = custom_encoder or {}
+ # Generated SDKs use Ellipsis (`...`) as the sentinel value for "OMIT".
+ # OMIT values should be excluded from serialized payloads.
+ if obj is Ellipsis:
+ return None
+ if custom_encoder:
+ if type(obj) in custom_encoder:
+ return custom_encoder[type(obj)](obj)
+ else:
+ for encoder_type, encoder_instance in custom_encoder.items():
+ if isinstance(obj, encoder_type):
+ return encoder_instance(obj)
+ if isinstance(obj, pydantic.BaseModel):
+ if IS_PYDANTIC_V2:
+ encoder = getattr(obj.model_config, "json_encoders", {}) # type: ignore # Pydantic v2
+ else:
+ encoder = getattr(obj.__config__, "json_encoders", {}) # type: ignore # Pydantic v1
+ if custom_encoder:
+ encoder.update(custom_encoder)
+ obj_dict = obj.dict(by_alias=True)
+ if "__root__" in obj_dict:
+ obj_dict = obj_dict["__root__"]
+ if "root" in obj_dict:
+ obj_dict = obj_dict["root"]
+ return jsonable_encoder(obj_dict, custom_encoder=encoder)
+ if dataclasses.is_dataclass(obj):
+ obj_dict = dataclasses.asdict(obj) # type: ignore
+ return jsonable_encoder(obj_dict, custom_encoder=custom_encoder)
+ if isinstance(obj, bytes):
+ return base64.b64encode(obj).decode("utf-8")
+ if isinstance(obj, Enum):
+ return obj.value
+ if isinstance(obj, PurePath):
+ return str(obj)
+ if isinstance(obj, (str, int, float, type(None))):
+ return obj
+ if isinstance(obj, dt.datetime):
+ return serialize_datetime(obj)
+ if isinstance(obj, dt.date):
+ return str(obj)
+ if isinstance(obj, dict):
+ encoded_dict = {}
+ allowed_keys = set(obj.keys())
+ for key, value in obj.items():
+ if key in allowed_keys:
+ if value is Ellipsis:
+ continue
+ encoded_key = jsonable_encoder(key, custom_encoder=custom_encoder)
+ encoded_value = jsonable_encoder(value, custom_encoder=custom_encoder)
+ encoded_dict[encoded_key] = encoded_value
+ return encoded_dict
+ if isinstance(obj, (list, set, frozenset, GeneratorType, tuple)):
+ encoded_list = []
+ for item in obj:
+ if item is Ellipsis:
+ continue
+ encoded_list.append(jsonable_encoder(item, custom_encoder=custom_encoder))
+ return encoded_list
+
+ def fallback_serializer(o: Any) -> Any:
+ attempt_encode = encode_by_type(o)
+ if attempt_encode is not None:
+ return attempt_encode
+
+ try:
+ data = dict(o)
+ except Exception as e:
+ errors: List[Exception] = []
+ errors.append(e)
+ try:
+ data = vars(o)
+ except Exception as e:
+ errors.append(e)
+ raise ValueError(errors) from e
+ return jsonable_encoder(data, custom_encoder=custom_encoder)
+
+ return to_jsonable_with_fallback(obj, fallback_serializer)
+
+
+def encode_path_param(obj: Any) -> str:
+ """Encode a value for use in a URL path segment.
+
+ Ensures proper string conversion for all types, including
+ booleans which need lowercase 'true'/'false' rather than
+ Python's 'True'/'False'.
+ """
+ if isinstance(obj, bool):
+ return "true" if obj else "false"
+ return str(jsonable_encoder(obj))
diff --git a/sdks/python/core/logging.py b/sdks/python/core/logging.py
new file mode 100644
index 0000000000..e5e572458b
--- /dev/null
+++ b/sdks/python/core/logging.py
@@ -0,0 +1,107 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import logging
+import typing
+
+LogLevel = typing.Literal["debug", "info", "warn", "error"]
+
+_LOG_LEVEL_MAP: typing.Dict[LogLevel, int] = {
+ "debug": 1,
+ "info": 2,
+ "warn": 3,
+ "error": 4,
+}
+
+
+class ILogger(typing.Protocol):
+ def debug(self, message: str, **kwargs: typing.Any) -> None: ...
+ def info(self, message: str, **kwargs: typing.Any) -> None: ...
+ def warn(self, message: str, **kwargs: typing.Any) -> None: ...
+ def error(self, message: str, **kwargs: typing.Any) -> None: ...
+
+
+class ConsoleLogger:
+ _logger: logging.Logger
+
+ def __init__(self) -> None:
+ self._logger = logging.getLogger("fern")
+ if not self._logger.handlers:
+ handler = logging.StreamHandler()
+ handler.setFormatter(logging.Formatter("%(levelname)s - %(message)s"))
+ self._logger.addHandler(handler)
+ self._logger.setLevel(logging.DEBUG)
+
+ def debug(self, message: str, **kwargs: typing.Any) -> None:
+ self._logger.debug(message, extra=kwargs)
+
+ def info(self, message: str, **kwargs: typing.Any) -> None:
+ self._logger.info(message, extra=kwargs)
+
+ def warn(self, message: str, **kwargs: typing.Any) -> None:
+ self._logger.warning(message, extra=kwargs)
+
+ def error(self, message: str, **kwargs: typing.Any) -> None:
+ self._logger.error(message, extra=kwargs)
+
+
+class LogConfig(typing.TypedDict, total=False):
+ level: LogLevel
+ logger: ILogger
+ silent: bool
+
+
+class Logger:
+ _level: int
+ _logger: ILogger
+ _silent: bool
+
+ def __init__(self, *, level: LogLevel, logger: ILogger, silent: bool) -> None:
+ self._level = _LOG_LEVEL_MAP[level]
+ self._logger = logger
+ self._silent = silent
+
+ def _should_log(self, level: LogLevel) -> bool:
+ return not self._silent and self._level <= _LOG_LEVEL_MAP[level]
+
+ def is_debug(self) -> bool:
+ return self._should_log("debug")
+
+ def is_info(self) -> bool:
+ return self._should_log("info")
+
+ def is_warn(self) -> bool:
+ return self._should_log("warn")
+
+ def is_error(self) -> bool:
+ return self._should_log("error")
+
+ def debug(self, message: str, **kwargs: typing.Any) -> None:
+ if self.is_debug():
+ self._logger.debug(message, **kwargs)
+
+ def info(self, message: str, **kwargs: typing.Any) -> None:
+ if self.is_info():
+ self._logger.info(message, **kwargs)
+
+ def warn(self, message: str, **kwargs: typing.Any) -> None:
+ if self.is_warn():
+ self._logger.warn(message, **kwargs)
+
+ def error(self, message: str, **kwargs: typing.Any) -> None:
+ if self.is_error():
+ self._logger.error(message, **kwargs)
+
+
+_default_logger: Logger = Logger(level="info", logger=ConsoleLogger(), silent=True)
+
+
+def create_logger(config: typing.Optional[typing.Union[LogConfig, Logger]] = None) -> Logger:
+ if config is None:
+ return _default_logger
+ if isinstance(config, Logger):
+ return config
+ return Logger(
+ level=config.get("level", "info"),
+ logger=config.get("logger", ConsoleLogger()),
+ silent=config.get("silent", True),
+ )
diff --git a/sdks/python/core/parse_error.py b/sdks/python/core/parse_error.py
new file mode 100644
index 0000000000..4527c6a8ad
--- /dev/null
+++ b/sdks/python/core/parse_error.py
@@ -0,0 +1,36 @@
+# This file was auto-generated by Fern from our API Definition.
+
+from typing import Any, Dict, Optional
+
+
+class ParsingError(Exception):
+ """
+ Raised when the SDK fails to parse/validate a response from the server.
+ This typically indicates that the server returned a response whose shape
+ does not match the expected schema.
+ """
+
+ headers: Optional[Dict[str, str]]
+ status_code: Optional[int]
+ body: Any
+ cause: Optional[Exception]
+
+ def __init__(
+ self,
+ *,
+ headers: Optional[Dict[str, str]] = None,
+ status_code: Optional[int] = None,
+ body: Any = None,
+ cause: Optional[Exception] = None,
+ ) -> None:
+ self.headers = headers
+ self.status_code = status_code
+ self.body = body
+ self.cause = cause
+ super().__init__()
+ if cause is not None:
+ self.__cause__ = cause
+
+ def __str__(self) -> str:
+ cause_str = f", cause: {self.cause}" if self.cause is not None else ""
+ return f"headers: {self.headers}, status_code: {self.status_code}, body: {self.body}{cause_str}"
diff --git a/sdks/python/core/pydantic_utilities.py b/sdks/python/core/pydantic_utilities.py
new file mode 100644
index 0000000000..6587f5e182
--- /dev/null
+++ b/sdks/python/core/pydantic_utilities.py
@@ -0,0 +1,508 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# nopycln: file
+import datetime as dt
+import inspect
+import json
+import logging
+from collections import defaultdict
+from dataclasses import asdict
+from typing import (
+ TYPE_CHECKING,
+ Any,
+ Callable,
+ ClassVar,
+ Dict,
+ List,
+ Mapping,
+ Optional,
+ Set,
+ Tuple,
+ Type,
+ TypeVar,
+ Union,
+ cast,
+)
+
+import pydantic
+import typing_extensions
+from pydantic.fields import FieldInfo as _FieldInfo
+
+_logger = logging.getLogger(__name__)
+
+if TYPE_CHECKING:
+ from .http_sse._models import ServerSentEvent
+
+IS_PYDANTIC_V2 = pydantic.VERSION.startswith("2.")
+
+if IS_PYDANTIC_V2:
+ _datetime_adapter = pydantic.TypeAdapter(dt.datetime) # type: ignore[attr-defined]
+ _date_adapter = pydantic.TypeAdapter(dt.date) # type: ignore[attr-defined]
+
+ def parse_datetime(value: Any) -> dt.datetime: # type: ignore[misc]
+ if isinstance(value, dt.datetime):
+ return value
+ return _datetime_adapter.validate_python(value)
+
+ def parse_date(value: Any) -> dt.date: # type: ignore[misc]
+ if isinstance(value, dt.datetime):
+ return value.date()
+ if isinstance(value, dt.date):
+ return value
+ return _date_adapter.validate_python(value)
+
+ # Avoid importing from pydantic.v1 to maintain Python 3.14 compatibility.
+ from typing import get_args as get_args # type: ignore[assignment]
+ from typing import get_origin as get_origin # type: ignore[assignment]
+
+ def is_literal_type(tp: Optional[Type[Any]]) -> bool: # type: ignore[misc]
+ return typing_extensions.get_origin(tp) is typing_extensions.Literal
+
+ def is_union(tp: Optional[Type[Any]]) -> bool: # type: ignore[misc]
+ return tp is Union or typing_extensions.get_origin(tp) is Union # type: ignore[comparison-overlap]
+
+ # Inline encoders_by_type to avoid importing from pydantic.v1.json
+ import re as _re
+ from collections import deque as _deque
+ from decimal import Decimal as _Decimal
+ from enum import Enum as _Enum
+ from ipaddress import (
+ IPv4Address as _IPv4Address,
+ )
+ from ipaddress import (
+ IPv4Interface as _IPv4Interface,
+ )
+ from ipaddress import (
+ IPv4Network as _IPv4Network,
+ )
+ from ipaddress import (
+ IPv6Address as _IPv6Address,
+ )
+ from ipaddress import (
+ IPv6Interface as _IPv6Interface,
+ )
+ from ipaddress import (
+ IPv6Network as _IPv6Network,
+ )
+ from pathlib import Path as _Path
+ from types import GeneratorType as _GeneratorType
+ from uuid import UUID as _UUID
+
+ from pydantic.fields import FieldInfo as ModelField # type: ignore[no-redef, assignment]
+
+ def _decimal_encoder(dec_value: Any) -> Any:
+ if dec_value.as_tuple().exponent >= 0:
+ return int(dec_value)
+ return float(dec_value)
+
+ encoders_by_type: Dict[Type[Any], Callable[[Any], Any]] = { # type: ignore[no-redef]
+ bytes: lambda o: o.decode(),
+ dt.date: lambda o: o.isoformat(),
+ dt.datetime: lambda o: o.isoformat(),
+ dt.time: lambda o: o.isoformat(),
+ dt.timedelta: lambda td: td.total_seconds(),
+ _Decimal: _decimal_encoder,
+ _Enum: lambda o: o.value,
+ frozenset: list,
+ _deque: list,
+ _GeneratorType: list,
+ _IPv4Address: str,
+ _IPv4Interface: str,
+ _IPv4Network: str,
+ _IPv6Address: str,
+ _IPv6Interface: str,
+ _IPv6Network: str,
+ _Path: str,
+ _re.Pattern: lambda o: o.pattern,
+ set: list,
+ _UUID: str,
+ }
+else:
+ from pydantic.datetime_parse import parse_date as parse_date # type: ignore[no-redef]
+ from pydantic.datetime_parse import parse_datetime as parse_datetime # type: ignore[no-redef]
+ from pydantic.fields import ModelField as ModelField # type: ignore[attr-defined, no-redef, assignment]
+ from pydantic.json import ENCODERS_BY_TYPE as encoders_by_type # type: ignore[no-redef]
+ from pydantic.typing import get_args as get_args # type: ignore[no-redef]
+ from pydantic.typing import get_origin as get_origin # type: ignore[no-redef]
+ from pydantic.typing import is_literal_type as is_literal_type # type: ignore[no-redef, assignment]
+ from pydantic.typing import is_union as is_union # type: ignore[no-redef]
+
+from .datetime_utils import serialize_datetime
+from .serialization import convert_and_respect_annotation_metadata
+from typing_extensions import TypeAlias
+
+T = TypeVar("T")
+Model = TypeVar("Model", bound=pydantic.BaseModel)
+
+
+def parse_sse_obj(sse: "ServerSentEvent", type_: Type[T]) -> T:
+ """
+ Parse a ServerSentEvent into the appropriate type.
+
+ This function handles data-level discrimination where the discriminator
+ (e.g., 'type') is inside the 'data' payload. It parses the SSE data field
+ as JSON and deserializes it into the target type.
+
+ Note: Protocol-level discrimination (where the discriminator comes from
+ the SSE event: field) is handled at code-generation time and does not
+ use this function.
+
+ Args:
+ sse: The ServerSentEvent object to parse
+ type_: The target type to deserialize into
+
+ Returns:
+ The parsed object of type T
+
+ Note:
+ This function is only available in SDK contexts where http_sse module exists.
+ """
+ sse_event = asdict(sse)
+ data_value = sse_event.get("data")
+ if isinstance(data_value, str) and data_value:
+ try:
+ parsed_data = json.loads(data_value)
+ return parse_obj_as(type_, parsed_data)
+ except json.JSONDecodeError as e:
+ _logger.warning(
+ "Failed to parse SSE data field as JSON: %s, data: %s",
+ e,
+ data_value[:100] if len(data_value) > 100 else data_value,
+ )
+ return parse_obj_as(type_, sse_event)
+
+
+_type_adapter_cache: Dict[int, Any] = {}
+
+
+def _get_type_adapter(type_: Type[Any]) -> Any:
+ key = id(type_)
+ adapter = _type_adapter_cache.get(key)
+ if adapter is None:
+ adapter = pydantic.TypeAdapter(type_) # type: ignore[attr-defined]
+ _type_adapter_cache[key] = adapter
+ return adapter
+
+
+def parse_obj_as(type_: Type[T], object_: Any) -> T:
+ # convert_and_respect_annotation_metadata is required for TypedDict aliasing.
+ #
+ # For Pydantic models, whether we should pre-dealias depends on how the model encodes aliasing:
+ # - If the model uses real Pydantic aliases (pydantic.Field(alias=...)), then we must pass wire keys through
+ # unchanged so Pydantic can validate them.
+ # - If the model encodes aliasing only via FieldMetadata annotations, then we MUST pre-dealias because Pydantic
+ # will not recognize those aliases during validation.
+ if inspect.isclass(type_) and issubclass(type_, pydantic.BaseModel):
+ has_pydantic_aliases = False
+ if IS_PYDANTIC_V2:
+ for field_name, field_info in getattr(type_, "model_fields", {}).items(): # type: ignore[attr-defined]
+ alias = getattr(field_info, "alias", None)
+ if alias is not None and alias != field_name:
+ has_pydantic_aliases = True
+ break
+ else:
+ for field in getattr(type_, "__fields__", {}).values():
+ alias = getattr(field, "alias", None)
+ name = getattr(field, "name", None)
+ if alias is not None and name is not None and alias != name:
+ has_pydantic_aliases = True
+ break
+
+ dealiased_object = (
+ object_
+ if has_pydantic_aliases
+ else convert_and_respect_annotation_metadata(object_=object_, annotation=type_, direction="read")
+ )
+ else:
+ dealiased_object = convert_and_respect_annotation_metadata(object_=object_, annotation=type_, direction="read")
+ if IS_PYDANTIC_V2:
+ adapter = _get_type_adapter(type_)
+ return adapter.validate_python(dealiased_object) # type: ignore[no-any-return]
+ return pydantic.parse_obj_as(type_, dealiased_object)
+
+
+def to_jsonable_with_fallback(obj: Any, fallback_serializer: Callable[[Any], Any]) -> Any:
+ if IS_PYDANTIC_V2:
+ from pydantic_core import to_jsonable_python
+
+ return to_jsonable_python(obj, fallback=fallback_serializer)
+ return fallback_serializer(obj)
+
+
+class UniversalBaseModel(pydantic.BaseModel):
+ if IS_PYDANTIC_V2:
+ model_config: ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict( # type: ignore[typeddict-unknown-key]
+ # Allow fields beginning with `model_` to be used in the model
+ protected_namespaces=(),
+ )
+
+ @pydantic.model_validator(mode="before") # type: ignore[attr-defined]
+ @classmethod
+ def _coerce_field_names_to_aliases(cls, data: Any) -> Any:
+ """
+ Accept Python field names in input by rewriting them to their Pydantic aliases,
+ while avoiding silent collisions when a key could refer to multiple fields.
+ """
+ if not isinstance(data, Mapping):
+ return data
+
+ fields = getattr(cls, "model_fields", {}) # type: ignore[attr-defined]
+ name_to_alias: Dict[str, str] = {}
+ alias_to_name: Dict[str, str] = {}
+
+ for name, field_info in fields.items():
+ alias = getattr(field_info, "alias", None) or name
+ name_to_alias[name] = alias
+ if alias != name:
+ alias_to_name[alias] = name
+
+ # Detect ambiguous keys: a key that is an alias for one field and a name for another.
+ ambiguous_keys = set(alias_to_name.keys()).intersection(set(name_to_alias.keys()))
+ for key in ambiguous_keys:
+ if key in data and name_to_alias[key] not in data:
+ raise ValueError(
+ f"Ambiguous input key '{key}': it is both a field name and an alias. "
+ "Provide the explicit alias key to disambiguate."
+ )
+
+ original_keys = set(data.keys())
+ rewritten: Dict[str, Any] = dict(data)
+ for name, alias in name_to_alias.items():
+ if alias != name and name in original_keys and alias not in rewritten:
+ rewritten[alias] = rewritten.pop(name)
+
+ return rewritten
+
+ @pydantic.model_serializer(mode="plain", when_used="json") # type: ignore[attr-defined]
+ def serialize_model(self) -> Any: # type: ignore[name-defined]
+ serialized = self.dict() # type: ignore[attr-defined]
+ data = {k: serialize_datetime(v) if isinstance(v, dt.datetime) else v for k, v in serialized.items()}
+ return data
+
+ else:
+
+ class Config:
+ smart_union = True
+ json_encoders = {dt.datetime: serialize_datetime}
+
+ @pydantic.root_validator(pre=True)
+ def _coerce_field_names_to_aliases(cls, values: Any) -> Any:
+ """
+ Pydantic v1 equivalent of _coerce_field_names_to_aliases.
+ """
+ if not isinstance(values, Mapping):
+ return values
+
+ fields = getattr(cls, "__fields__", {})
+ name_to_alias: Dict[str, str] = {}
+ alias_to_name: Dict[str, str] = {}
+
+ for name, field in fields.items():
+ alias = getattr(field, "alias", None) or name
+ name_to_alias[name] = alias
+ if alias != name:
+ alias_to_name[alias] = name
+
+ ambiguous_keys = set(alias_to_name.keys()).intersection(set(name_to_alias.keys()))
+ for key in ambiguous_keys:
+ if key in values and name_to_alias[key] not in values:
+ raise ValueError(
+ f"Ambiguous input key '{key}': it is both a field name and an alias. "
+ "Provide the explicit alias key to disambiguate."
+ )
+
+ original_keys = set(values.keys())
+ rewritten: Dict[str, Any] = dict(values)
+ for name, alias in name_to_alias.items():
+ if alias != name and name in original_keys and alias not in rewritten:
+ rewritten[alias] = rewritten.pop(name)
+
+ return rewritten
+
+ @classmethod
+ def model_construct(cls: Type["Model"], _fields_set: Optional[Set[str]] = None, **values: Any) -> "Model":
+ dealiased_object = convert_and_respect_annotation_metadata(object_=values, annotation=cls, direction="read")
+ return cls.construct(_fields_set, **dealiased_object)
+
+ @classmethod
+ def construct(cls: Type["Model"], _fields_set: Optional[Set[str]] = None, **values: Any) -> "Model":
+ dealiased_object = convert_and_respect_annotation_metadata(object_=values, annotation=cls, direction="read")
+ if IS_PYDANTIC_V2:
+ return super().model_construct(_fields_set, **dealiased_object) # type: ignore[misc]
+ return super().construct(_fields_set, **dealiased_object)
+
+ def json(self, **kwargs: Any) -> str:
+ kwargs_with_defaults = {
+ "by_alias": True,
+ "exclude_unset": True,
+ **kwargs,
+ }
+ if IS_PYDANTIC_V2:
+ return super().model_dump_json(**kwargs_with_defaults) # type: ignore[misc]
+ return super().json(**kwargs_with_defaults)
+
+ def dict(self, **kwargs: Any) -> Dict[str, Any]:
+ """
+ Override the default dict method to `exclude_unset` by default. This function patches
+ `exclude_unset` to work include fields within non-None default values.
+ """
+ # Note: the logic here is multiplexed given the levers exposed in Pydantic V1 vs V2
+ # Pydantic V1's .dict can be extremely slow, so we do not want to call it twice.
+ #
+ # We'd ideally do the same for Pydantic V2, but it shells out to a library to serialize models
+ # that we have less control over, and this is less intrusive than custom serializers for now.
+ if IS_PYDANTIC_V2:
+ kwargs_with_defaults_exclude_unset = {
+ **kwargs,
+ "by_alias": True,
+ "exclude_unset": True,
+ "exclude_none": False,
+ }
+ kwargs_with_defaults_exclude_none = {
+ **kwargs,
+ "by_alias": True,
+ "exclude_none": True,
+ "exclude_unset": False,
+ }
+ dict_dump = deep_union_pydantic_dicts(
+ super().model_dump(**kwargs_with_defaults_exclude_unset), # type: ignore[misc]
+ super().model_dump(**kwargs_with_defaults_exclude_none), # type: ignore[misc]
+ )
+
+ else:
+ _fields_set = self.__fields_set__.copy()
+
+ fields = _get_model_fields(self.__class__)
+ for name, field in fields.items():
+ if name not in _fields_set:
+ default = _get_field_default(field)
+
+ # If the default values are non-null act like they've been set
+ # This effectively allows exclude_unset to work like exclude_none where
+ # the latter passes through intentionally set none values.
+ if default is not None or ("exclude_unset" in kwargs and not kwargs["exclude_unset"]):
+ _fields_set.add(name)
+
+ if default is not None:
+ self.__fields_set__.add(name)
+
+ kwargs_with_defaults_exclude_unset_include_fields = {
+ "by_alias": True,
+ "exclude_unset": True,
+ "include": _fields_set,
+ **kwargs,
+ }
+
+ dict_dump = super().dict(**kwargs_with_defaults_exclude_unset_include_fields)
+
+ return cast(
+ Dict[str, Any],
+ convert_and_respect_annotation_metadata(object_=dict_dump, annotation=self.__class__, direction="write"),
+ )
+
+
+def _union_list_of_pydantic_dicts(source: List[Any], destination: List[Any]) -> List[Any]:
+ converted_list: List[Any] = []
+ for i, item in enumerate(source):
+ destination_value = destination[i]
+ if isinstance(item, dict):
+ converted_list.append(deep_union_pydantic_dicts(item, destination_value))
+ elif isinstance(item, list):
+ converted_list.append(_union_list_of_pydantic_dicts(item, destination_value))
+ else:
+ converted_list.append(item)
+ return converted_list
+
+
+def deep_union_pydantic_dicts(source: Dict[str, Any], destination: Dict[str, Any]) -> Dict[str, Any]:
+ for key, value in source.items():
+ node = destination.setdefault(key, {})
+ if isinstance(value, dict):
+ deep_union_pydantic_dicts(value, node)
+ # Note: we do not do this same processing for sets given we do not have sets of models
+ # and given the sets are unordered, the processing of the set and matching objects would
+ # be non-trivial.
+ elif isinstance(value, list):
+ destination[key] = _union_list_of_pydantic_dicts(value, node)
+ else:
+ destination[key] = value
+
+ return destination
+
+
+if IS_PYDANTIC_V2:
+
+ class V2RootModel(UniversalBaseModel, pydantic.RootModel): # type: ignore[misc, name-defined, type-arg]
+ pass
+
+ UniversalRootModel: TypeAlias = V2RootModel # type: ignore[misc]
+else:
+ UniversalRootModel: TypeAlias = UniversalBaseModel # type: ignore[misc, no-redef]
+
+
+def encode_by_type(o: Any) -> Any:
+ encoders_by_class_tuples: Dict[Callable[[Any], Any], Tuple[Any, ...]] = defaultdict(tuple)
+ for type_, encoder in encoders_by_type.items():
+ encoders_by_class_tuples[encoder] += (type_,)
+
+ if type(o) in encoders_by_type:
+ return encoders_by_type[type(o)](o)
+ for encoder, classes_tuple in encoders_by_class_tuples.items():
+ if isinstance(o, classes_tuple):
+ return encoder(o)
+
+
+def update_forward_refs(model: Type["Model"], **localns: Any) -> None:
+ if IS_PYDANTIC_V2:
+ model.model_rebuild(raise_errors=False) # type: ignore[attr-defined]
+ else:
+ model.update_forward_refs(**localns)
+
+
+# Mirrors Pydantic's internal typing
+AnyCallable = Callable[..., Any]
+
+
+def universal_root_validator(
+ pre: bool = False,
+) -> Callable[[AnyCallable], AnyCallable]:
+ def decorator(func: AnyCallable) -> AnyCallable:
+ if IS_PYDANTIC_V2:
+ # In Pydantic v2, for RootModel we always use "before" mode
+ # The custom validators transform the input value before the model is created
+ return cast(AnyCallable, pydantic.model_validator(mode="before")(func)) # type: ignore[attr-defined]
+ return cast(AnyCallable, pydantic.root_validator(pre=pre)(func)) # type: ignore[call-overload]
+
+ return decorator
+
+
+def universal_field_validator(field_name: str, pre: bool = False) -> Callable[[AnyCallable], AnyCallable]:
+ def decorator(func: AnyCallable) -> AnyCallable:
+ if IS_PYDANTIC_V2:
+ return cast(AnyCallable, pydantic.field_validator(field_name, mode="before" if pre else "after")(func)) # type: ignore[attr-defined]
+ return cast(AnyCallable, pydantic.validator(field_name, pre=pre)(func))
+
+ return decorator
+
+
+PydanticField = Union[ModelField, _FieldInfo]
+
+
+def _get_model_fields(model: Type["Model"]) -> Mapping[str, PydanticField]:
+ if IS_PYDANTIC_V2:
+ return cast(Mapping[str, PydanticField], model.model_fields) # type: ignore[attr-defined]
+ return cast(Mapping[str, PydanticField], model.__fields__)
+
+
+def _get_field_default(field: PydanticField) -> Any:
+ try:
+ value = field.get_default() # type: ignore[union-attr]
+ except:
+ value = field.default
+ if IS_PYDANTIC_V2:
+ from pydantic_core import PydanticUndefined
+
+ if value == PydanticUndefined:
+ return None
+ return value
+ return value
diff --git a/sdks/python/core/query_encoder.py b/sdks/python/core/query_encoder.py
new file mode 100644
index 0000000000..3183001d40
--- /dev/null
+++ b/sdks/python/core/query_encoder.py
@@ -0,0 +1,58 @@
+# This file was auto-generated by Fern from our API Definition.
+
+from typing import Any, Dict, List, Optional, Tuple
+
+import pydantic
+
+
+# Flattens dicts to be of the form {"key[subkey][subkey2]": value} where value is not a dict
+def traverse_query_dict(dict_flat: Dict[str, Any], key_prefix: Optional[str] = None) -> List[Tuple[str, Any]]:
+ result = []
+ for k, v in dict_flat.items():
+ key = f"{key_prefix}[{k}]" if key_prefix is not None else k
+ if isinstance(v, dict):
+ result.extend(traverse_query_dict(v, key))
+ elif isinstance(v, list):
+ for arr_v in v:
+ if isinstance(arr_v, dict):
+ result.extend(traverse_query_dict(arr_v, key))
+ else:
+ result.append((key, arr_v))
+ else:
+ result.append((key, v))
+ return result
+
+
+def single_query_encoder(query_key: str, query_value: Any) -> List[Tuple[str, Any]]:
+ if isinstance(query_value, pydantic.BaseModel) or isinstance(query_value, dict):
+ if isinstance(query_value, pydantic.BaseModel):
+ obj_dict = query_value.dict(by_alias=True)
+ else:
+ obj_dict = query_value
+ return traverse_query_dict(obj_dict, query_key)
+ elif isinstance(query_value, list):
+ encoded_values: List[Tuple[str, Any]] = []
+ for value in query_value:
+ if isinstance(value, pydantic.BaseModel) or isinstance(value, dict):
+ if isinstance(value, pydantic.BaseModel):
+ obj_dict = value.dict(by_alias=True)
+ elif isinstance(value, dict):
+ obj_dict = value
+
+ encoded_values.extend(single_query_encoder(query_key, obj_dict))
+ else:
+ encoded_values.append((query_key, value))
+
+ return encoded_values
+
+ return [(query_key, query_value)]
+
+
+def encode_query(query: Optional[Dict[str, Any]]) -> Optional[List[Tuple[str, Any]]]:
+ if query is None:
+ return None
+
+ encoded_query = []
+ for k, v in query.items():
+ encoded_query.extend(single_query_encoder(k, v))
+ return encoded_query
diff --git a/sdks/python/core/remove_none_from_dict.py b/sdks/python/core/remove_none_from_dict.py
new file mode 100644
index 0000000000..c2298143f1
--- /dev/null
+++ b/sdks/python/core/remove_none_from_dict.py
@@ -0,0 +1,11 @@
+# This file was auto-generated by Fern from our API Definition.
+
+from typing import Any, Dict, Mapping, Optional
+
+
+def remove_none_from_dict(original: Mapping[str, Optional[Any]]) -> Dict[str, Any]:
+ new: Dict[str, Any] = {}
+ for key, value in original.items():
+ if value is not None:
+ new[key] = value
+ return new
diff --git a/sdks/python/core/request_options.py b/sdks/python/core/request_options.py
new file mode 100644
index 0000000000..1b38804432
--- /dev/null
+++ b/sdks/python/core/request_options.py
@@ -0,0 +1,35 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+try:
+ from typing import NotRequired # type: ignore
+except ImportError:
+ from typing_extensions import NotRequired
+
+
+class RequestOptions(typing.TypedDict, total=False):
+ """
+ Additional options for request-specific configuration when calling APIs via the SDK.
+ This is used primarily as an optional final parameter for service functions.
+
+ Attributes:
+ - timeout_in_seconds: int. The number of seconds to await an API call before timing out.
+
+ - max_retries: int. The max number of retries to attempt if the API call fails.
+
+ - additional_headers: typing.Dict[str, typing.Any]. A dictionary containing additional parameters to spread into the request's header dict
+
+ - additional_query_parameters: typing.Dict[str, typing.Any]. A dictionary containing additional parameters to spread into the request's query parameters dict
+
+ - additional_body_parameters: typing.Dict[str, typing.Any]. A dictionary containing additional parameters to spread into the request's body parameters dict
+
+ - chunk_size: int. The size, in bytes, to process each chunk of data being streamed back within the response. This equates to leveraging `chunk_size` within `requests` or `httpx`, and is only leveraged for file downloads.
+ """
+
+ timeout_in_seconds: NotRequired[int]
+ max_retries: NotRequired[int]
+ additional_headers: NotRequired[typing.Dict[str, typing.Any]]
+ additional_query_parameters: NotRequired[typing.Dict[str, typing.Any]]
+ additional_body_parameters: NotRequired[typing.Dict[str, typing.Any]]
+ chunk_size: NotRequired[int]
diff --git a/sdks/python/core/serialization.py b/sdks/python/core/serialization.py
new file mode 100644
index 0000000000..1d753e26f7
--- /dev/null
+++ b/sdks/python/core/serialization.py
@@ -0,0 +1,347 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import collections
+import inspect
+import typing
+
+import pydantic
+import typing_extensions
+
+
+class FieldMetadata:
+ """
+ Metadata class used to annotate fields to provide additional information.
+
+ Example:
+ class MyDict(TypedDict):
+ field: typing.Annotated[str, FieldMetadata(alias="field_name")]
+
+ Will serialize: `{"field": "value"}`
+ To: `{"field_name": "value"}`
+ """
+
+ alias: str
+
+ def __init__(self, *, alias: str) -> None:
+ self.alias = alias
+
+
+# Resolving type hints (typing.get_type_hints) is expensive because it eval/compiles
+# forward-reference annotations. The result is constant for a given type, so we cache it.
+# This is critical for hot paths like SSE event parsing, where the same (often large
+# discriminated-union) type is converted on every single event.
+_type_hints_cache: typing.Dict[typing.Any, typing.Dict[str, typing.Any]] = {}
+
+
+def _get_cached_type_hints(expected_type: typing.Any) -> typing.Dict[str, typing.Any]:
+ try:
+ cached = _type_hints_cache.get(expected_type)
+ except TypeError:
+ # Unhashable type; resolve without caching.
+ return _resolve_type_hints(expected_type)
+ if cached is None:
+ cached = _resolve_type_hints(expected_type)
+ _type_hints_cache[expected_type] = cached
+ return cached
+
+
+def _resolve_type_hints(expected_type: typing.Any) -> typing.Dict[str, typing.Any]:
+ try:
+ return typing_extensions.get_type_hints(expected_type, include_extras=True)
+ except NameError:
+ # The type contains a circular reference, so we use the __annotations__ attribute directly.
+ return getattr(expected_type, "__annotations__", {})
+
+
+# Whether convert_and_respect_annotation_metadata can possibly rewrite anything for a given
+# annotation, i.e. whether any reachable model/TypedDict field carries a FieldMetadata alias.
+# This is constant per type, so we cache it and use it to short-circuit the recursive walk.
+_requires_conversion_cache: typing.Dict[typing.Any, bool] = {}
+
+
+def _requires_conversion(type_: typing.Any) -> bool:
+ try:
+ cached = _requires_conversion_cache.get(type_)
+ except TypeError:
+ # Unhashable annotation; compute without caching.
+ return _compute_requires_conversion(type_, set())
+ if cached is None:
+ cached = _compute_requires_conversion(type_, set())
+ _requires_conversion_cache[type_] = cached
+ return cached
+
+
+def _compute_requires_conversion(type_: typing.Any, seen: typing.Set[typing.Any]) -> bool:
+ clean_type = _remove_annotations(type_)
+
+ try:
+ if clean_type in seen:
+ return False
+ seen = seen | {clean_type}
+ except TypeError:
+ # Unhashable type; skip cycle tracking (the type graph is finite in practice).
+ pass
+
+ # Models / TypedDicts: a field alias here means we must dealias; otherwise recurse into fields.
+ if (inspect.isclass(clean_type) and issubclass(clean_type, pydantic.BaseModel)) or typing_extensions.is_typeddict(
+ clean_type
+ ):
+ annotations = _get_cached_type_hints(clean_type)
+ if _get_alias_to_field_name(annotations):
+ return True
+ return any(_compute_requires_conversion(hint, seen) for hint in annotations.values())
+
+ # Containers / unions: recurse into the type arguments (List/Set/Sequence/Dict/Union/etc.).
+ return any(_compute_requires_conversion(arg, seen) for arg in typing_extensions.get_args(clean_type))
+
+
+def convert_and_respect_annotation_metadata(
+ *,
+ object_: typing.Any,
+ annotation: typing.Any,
+ inner_type: typing.Optional[typing.Any] = None,
+ direction: typing.Literal["read", "write"],
+) -> typing.Any:
+ """
+ Respect the metadata annotations on a field, such as aliasing. This function effectively
+ manipulates the dict-form of an object to respect the metadata annotations. This is primarily used for
+ TypedDicts, which cannot support aliasing out of the box, and can be extended for additional
+ utilities, such as defaults.
+
+ Parameters
+ ----------
+ object_ : typing.Any
+
+ annotation : type
+ The type we're looking to apply typing annotations from
+
+ inner_type : typing.Optional[type]
+
+ Returns
+ -------
+ typing.Any
+ """
+
+ if object_ is None:
+ return None
+ if inner_type is None:
+ inner_type = annotation
+ # The only thing this function ever rewrites is keys that carry a FieldMetadata
+ # alias. If nothing in the (cached) type graph has such an alias, the conversion is
+ # a content-identity transform, so we can skip the entire recursive walk. This is
+ # the hot path for SSE streaming, where a large discriminated union would otherwise
+ # be traversed on every single event.
+ if not _requires_conversion(annotation):
+ return object_
+
+ clean_type = _remove_annotations(inner_type)
+ # Pydantic models
+ if (
+ inspect.isclass(clean_type)
+ and issubclass(clean_type, pydantic.BaseModel)
+ and isinstance(object_, typing.Mapping)
+ ):
+ return _convert_mapping(object_, clean_type, direction)
+ # TypedDicts
+ if typing_extensions.is_typeddict(clean_type) and isinstance(object_, typing.Mapping):
+ return _convert_mapping(object_, clean_type, direction)
+
+ if (
+ typing_extensions.get_origin(clean_type) == typing.Dict
+ or typing_extensions.get_origin(clean_type) == dict
+ or clean_type == typing.Dict
+ ) and isinstance(object_, typing.Dict):
+ key_type = typing_extensions.get_args(clean_type)[0]
+ value_type = typing_extensions.get_args(clean_type)[1]
+
+ return {
+ key: convert_and_respect_annotation_metadata(
+ object_=value,
+ annotation=annotation,
+ inner_type=value_type,
+ direction=direction,
+ )
+ for key, value in object_.items()
+ }
+
+ # If you're iterating on a string, do not bother to coerce it to a sequence.
+ if not isinstance(object_, str):
+ if (
+ typing_extensions.get_origin(clean_type) == typing.Set
+ or typing_extensions.get_origin(clean_type) == set
+ or clean_type == typing.Set
+ ) and isinstance(object_, typing.Set):
+ inner_type = typing_extensions.get_args(clean_type)[0]
+ return {
+ convert_and_respect_annotation_metadata(
+ object_=item,
+ annotation=annotation,
+ inner_type=inner_type,
+ direction=direction,
+ )
+ for item in object_
+ }
+ elif (
+ (
+ typing_extensions.get_origin(clean_type) == typing.List
+ or typing_extensions.get_origin(clean_type) == list
+ or clean_type == typing.List
+ )
+ and isinstance(object_, typing.List)
+ ) or (
+ (
+ typing_extensions.get_origin(clean_type) == typing.Sequence
+ or typing_extensions.get_origin(clean_type) == collections.abc.Sequence
+ or clean_type == typing.Sequence
+ )
+ and isinstance(object_, typing.Sequence)
+ ):
+ inner_type = typing_extensions.get_args(clean_type)[0]
+ return [
+ convert_and_respect_annotation_metadata(
+ object_=item,
+ annotation=annotation,
+ inner_type=inner_type,
+ direction=direction,
+ )
+ for item in object_
+ ]
+
+ if typing_extensions.get_origin(clean_type) == typing.Union:
+ # We should be able to ~relatively~ safely try to convert keys against all
+ # member types in the union, the edge case here is if one member aliases a field
+ # of the same name to a different name from another member
+ # Or if another member aliases a field of the same name that another member does not.
+ for member in typing_extensions.get_args(clean_type):
+ object_ = convert_and_respect_annotation_metadata(
+ object_=object_,
+ annotation=annotation,
+ inner_type=member,
+ direction=direction,
+ )
+ return object_
+
+ annotated_type = _get_annotation(annotation)
+ if annotated_type is None:
+ return object_
+
+ # If the object is not a TypedDict, a Union, or other container (list, set, sequence, etc.)
+ # Then we can safely call it on the recursive conversion.
+ return object_
+
+
+def _convert_mapping(
+ object_: typing.Mapping[str, object],
+ expected_type: typing.Any,
+ direction: typing.Literal["read", "write"],
+) -> typing.Mapping[str, object]:
+ converted_object: typing.Dict[str, object] = {}
+ annotations = _get_cached_type_hints(expected_type)
+ aliases_to_field_names = _get_alias_to_field_name(annotations)
+ for key, value in object_.items():
+ if direction == "read" and key in aliases_to_field_names:
+ dealiased_key = aliases_to_field_names.get(key)
+ if dealiased_key is not None:
+ type_ = annotations.get(dealiased_key)
+ else:
+ type_ = annotations.get(key)
+ # Note you can't get the annotation by the field name if you're in read mode, so you must check the aliases map
+ #
+ # So this is effectively saying if we're in write mode, and we don't have a type, or if we're in read mode and we don't have an alias
+ # then we can just pass the value through as is
+ if type_ is None:
+ converted_object[key] = value
+ elif direction == "read" and key not in aliases_to_field_names:
+ converted_object[key] = convert_and_respect_annotation_metadata(
+ object_=value, annotation=type_, direction=direction
+ )
+ else:
+ converted_object[_alias_key(key, type_, direction, aliases_to_field_names)] = (
+ convert_and_respect_annotation_metadata(object_=value, annotation=type_, direction=direction)
+ )
+ return converted_object
+
+
+def _get_annotation(type_: typing.Any) -> typing.Optional[typing.Any]:
+ maybe_annotated_type = typing_extensions.get_origin(type_)
+ if maybe_annotated_type is None:
+ return None
+
+ if maybe_annotated_type == typing_extensions.NotRequired:
+ type_ = typing_extensions.get_args(type_)[0]
+ maybe_annotated_type = typing_extensions.get_origin(type_)
+
+ if maybe_annotated_type == typing_extensions.Annotated:
+ return type_
+
+ return None
+
+
+def _remove_annotations(type_: typing.Any) -> typing.Any:
+ maybe_annotated_type = typing_extensions.get_origin(type_)
+ if maybe_annotated_type is None:
+ return type_
+
+ if maybe_annotated_type == typing_extensions.NotRequired:
+ return _remove_annotations(typing_extensions.get_args(type_)[0])
+
+ if maybe_annotated_type == typing_extensions.Annotated:
+ return _remove_annotations(typing_extensions.get_args(type_)[0])
+
+ return type_
+
+
+def get_alias_to_field_mapping(type_: typing.Any) -> typing.Dict[str, str]:
+ annotations = _get_cached_type_hints(type_)
+ return _get_alias_to_field_name(annotations)
+
+
+def get_field_to_alias_mapping(type_: typing.Any) -> typing.Dict[str, str]:
+ annotations = _get_cached_type_hints(type_)
+ return _get_field_to_alias_name(annotations)
+
+
+def _get_alias_to_field_name(
+ field_to_hint: typing.Dict[str, typing.Any],
+) -> typing.Dict[str, str]:
+ aliases = {}
+ for field, hint in field_to_hint.items():
+ maybe_alias = _get_alias_from_type(hint)
+ if maybe_alias is not None:
+ aliases[maybe_alias] = field
+ return aliases
+
+
+def _get_field_to_alias_name(
+ field_to_hint: typing.Dict[str, typing.Any],
+) -> typing.Dict[str, str]:
+ aliases = {}
+ for field, hint in field_to_hint.items():
+ maybe_alias = _get_alias_from_type(hint)
+ if maybe_alias is not None:
+ aliases[field] = maybe_alias
+ return aliases
+
+
+def _get_alias_from_type(type_: typing.Any) -> typing.Optional[str]:
+ maybe_annotated_type = _get_annotation(type_)
+
+ if maybe_annotated_type is not None:
+ # The actual annotations are 1 onward, the first is the annotated type
+ annotations = typing_extensions.get_args(maybe_annotated_type)[1:]
+
+ for annotation in annotations:
+ if isinstance(annotation, FieldMetadata) and annotation.alias is not None:
+ return annotation.alias
+ return None
+
+
+def _alias_key(
+ key: str,
+ type_: typing.Any,
+ direction: typing.Literal["read", "write"],
+ aliases_to_field_names: typing.Dict[str, str],
+) -> str:
+ if direction == "read":
+ return aliases_to_field_names.get(key, key)
+ return _get_alias_from_type(type_=type_) or key
diff --git a/sdks/python/discovery/__init__.py b/sdks/python/discovery/__init__.py
new file mode 100644
index 0000000000..5cde0202dc
--- /dev/null
+++ b/sdks/python/discovery/__init__.py
@@ -0,0 +1,4 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
diff --git a/sdks/python/discovery/client.py b/sdks/python/discovery/client.py
new file mode 100644
index 0000000000..ed515d9857
--- /dev/null
+++ b/sdks/python/discovery/client.py
@@ -0,0 +1,134 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.request_options import RequestOptions
+from ..types.auth_discovery_response import AuthDiscoveryResponse
+from .raw_client import AsyncRawDiscoveryClient, RawDiscoveryClient
+
+
+class DiscoveryClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._raw_client = RawDiscoveryClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> RawDiscoveryClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ RawDiscoveryClient
+ """
+ return self._raw_client
+
+ def get_auth_discovery(self, *, request_options: typing.Optional[RequestOptions] = None) -> AuthDiscoveryResponse:
+ """
+ Return authentication configuration for CLI/SDK discovery.
+
+ This endpoint is unauthenticated and returns the information clients
+ need to authenticate with this NeMo Platform deployment.
+
+ **Response fields:**
+
+ - `auth_enabled`: Whether authentication is enabled on this cluster
+ - `oidc`: OIDC configuration (only present when OIDC is enabled)
+ - `issuer`: The OIDC issuer URL
+ - `authorization_endpoint`: Authorization endpoint for browser-based flows
+ - `token_endpoint`: Token exchange endpoint
+ - `device_authorization_endpoint`: Device flow authorization endpoint (for CLI)
+ - `userinfo_endpoint`: UserInfo endpoint
+ - `client_id`: OAuth client ID to use
+ - `default_scopes`: OAuth scopes to request during authentication
+ - `scope_prefix`: Prefix to prepend to custom scopes (those with ':' or '.default')
+
+ Parameters
+ ----------
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AuthDiscoveryResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.discovery.get_auth_discovery()
+ """
+ _response = self._raw_client.get_auth_discovery(request_options=request_options)
+ return _response.data
+
+
+class AsyncDiscoveryClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._raw_client = AsyncRawDiscoveryClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> AsyncRawDiscoveryClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ AsyncRawDiscoveryClient
+ """
+ return self._raw_client
+
+ async def get_auth_discovery(
+ self, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AuthDiscoveryResponse:
+ """
+ Return authentication configuration for CLI/SDK discovery.
+
+ This endpoint is unauthenticated and returns the information clients
+ need to authenticate with this NeMo Platform deployment.
+
+ **Response fields:**
+
+ - `auth_enabled`: Whether authentication is enabled on this cluster
+ - `oidc`: OIDC configuration (only present when OIDC is enabled)
+ - `issuer`: The OIDC issuer URL
+ - `authorization_endpoint`: Authorization endpoint for browser-based flows
+ - `token_endpoint`: Token exchange endpoint
+ - `device_authorization_endpoint`: Device flow authorization endpoint (for CLI)
+ - `userinfo_endpoint`: UserInfo endpoint
+ - `client_id`: OAuth client ID to use
+ - `default_scopes`: OAuth scopes to request during authentication
+ - `scope_prefix`: Prefix to prepend to custom scopes (those with ':' or '.default')
+
+ Parameters
+ ----------
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AuthDiscoveryResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.discovery.get_auth_discovery()
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_auth_discovery(request_options=request_options)
+ return _response.data
diff --git a/sdks/python/discovery/raw_client.py b/sdks/python/discovery/raw_client.py
new file mode 100644
index 0000000000..244863f8e0
--- /dev/null
+++ b/sdks/python/discovery/raw_client.py
@@ -0,0 +1,135 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+from json.decoder import JSONDecodeError
+
+from ..core.api_error import ApiError
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.http_response import AsyncHttpResponse, HttpResponse
+from ..core.parse_error import ParsingError
+from ..core.pydantic_utilities import parse_obj_as
+from ..core.request_options import RequestOptions
+from ..types.auth_discovery_response import AuthDiscoveryResponse
+from pydantic import ValidationError
+
+
+class RawDiscoveryClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ def get_auth_discovery(
+ self, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[AuthDiscoveryResponse]:
+ """
+ Return authentication configuration for CLI/SDK discovery.
+
+ This endpoint is unauthenticated and returns the information clients
+ need to authenticate with this NeMo Platform deployment.
+
+ **Response fields:**
+
+ - `auth_enabled`: Whether authentication is enabled on this cluster
+ - `oidc`: OIDC configuration (only present when OIDC is enabled)
+ - `issuer`: The OIDC issuer URL
+ - `authorization_endpoint`: Authorization endpoint for browser-based flows
+ - `token_endpoint`: Token exchange endpoint
+ - `device_authorization_endpoint`: Device flow authorization endpoint (for CLI)
+ - `userinfo_endpoint`: UserInfo endpoint
+ - `client_id`: OAuth client ID to use
+ - `default_scopes`: OAuth scopes to request during authentication
+ - `scope_prefix`: Prefix to prepend to custom scopes (those with ':' or '.default')
+
+ Parameters
+ ----------
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[AuthDiscoveryResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ "apis/auth/discovery",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ AuthDiscoveryResponse,
+ parse_obj_as(
+ type_=AuthDiscoveryResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+
+class AsyncRawDiscoveryClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ async def get_auth_discovery(
+ self, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[AuthDiscoveryResponse]:
+ """
+ Return authentication configuration for CLI/SDK discovery.
+
+ This endpoint is unauthenticated and returns the information clients
+ need to authenticate with this NeMo Platform deployment.
+
+ **Response fields:**
+
+ - `auth_enabled`: Whether authentication is enabled on this cluster
+ - `oidc`: OIDC configuration (only present when OIDC is enabled)
+ - `issuer`: The OIDC issuer URL
+ - `authorization_endpoint`: Authorization endpoint for browser-based flows
+ - `token_endpoint`: Token exchange endpoint
+ - `device_authorization_endpoint`: Device flow authorization endpoint (for CLI)
+ - `userinfo_endpoint`: UserInfo endpoint
+ - `client_id`: OAuth client ID to use
+ - `default_scopes`: OAuth scopes to request during authentication
+ - `scope_prefix`: Prefix to prepend to custom scopes (those with ':' or '.default')
+
+ Parameters
+ ----------
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[AuthDiscoveryResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ "apis/auth/discovery",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ AuthDiscoveryResponse,
+ parse_obj_as(
+ type_=AuthDiscoveryResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
diff --git a/sdks/python/entity_store/__init__.py b/sdks/python/entity_store/__init__.py
new file mode 100644
index 0000000000..5cde0202dc
--- /dev/null
+++ b/sdks/python/entity_store/__init__.py
@@ -0,0 +1,4 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
diff --git a/sdks/python/entity_store/client.py b/sdks/python/entity_store/client.py
new file mode 100644
index 0000000000..030e30c73d
--- /dev/null
+++ b/sdks/python/entity_store/client.py
@@ -0,0 +1,2457 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.request_options import RequestOptions
+from ..types.delete_response import DeleteResponse
+from ..types.entities_page import EntitiesPage
+from ..types.entity import Entity
+from ..types.generic_sort_field import GenericSortField
+from ..types.project import Project
+from ..types.project_sort_field import ProjectSortField
+from ..types.projects_page import ProjectsPage
+from ..types.workspace import Workspace
+from ..types.workspace_member import WorkspaceMember
+from ..types.workspace_member_list_response import WorkspaceMemberListResponse
+from ..types.workspaces_page import WorkspacesPage
+from .raw_client import AsyncRawEntityStoreClient, RawEntityStoreClient
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class EntityStoreClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._raw_client = RawEntityStoreClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> RawEntityStoreClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ RawEntityStoreClient
+ """
+ return self._raw_client
+
+ def get_entity_by_id(self, id: str, *, request_options: typing.Optional[RequestOptions] = None) -> Entity:
+ """
+ Get a specific entity by its unique identifier.
+ This endpoint is primarily for debugging and internal use.
+
+ Example:
+ ```
+ GET /apis/entities/v2/entities/customization-config-5Q2LoF8z8M9JZxZsHwJKNn
+ ```
+
+ Parameters
+ ----------
+ id : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Entity
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.entity_store.get_entity_by_id(
+ id="id",
+ )
+ """
+ _response = self._raw_client.get_entity_by_id(id, request_options=request_options)
+ return _response.data
+
+ def list_workspaces(
+ self,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[GenericSortField] = None,
+ filter: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> WorkspacesPage:
+ """
+ List all workspaces with pagination.
+
+ When authentication is enabled, only workspaces the principal has access to
+ are returned. Service principals and platform admins have access to all workspaces.
+
+ Query Parameters:
+ - page, page_size: Pagination
+ - sort: Sort field
+ - filter: Advanced filters (JSON, text, or bracket notation)
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces?sort=-created_at&page=1&page_size=10
+ ```
+
+ Parameters
+ ----------
+ page : typing.Optional[int]
+ Page number
+
+ page_size : typing.Optional[int]
+ Items per page
+
+ sort : typing.Optional[GenericSortField]
+ Sort field
+
+ filter : typing.Optional[str]
+ Query filter expression. Supports text and JSON syntaxes:
+ - Text: name:"value" AND status>500 with operators : ~ > >= < <= IN NOT IN AND OR and negation prefix -
+ - Object (JSON): {"name":{"$like":"value"}} with operators $eq, $like, $lt, $lte, $gt, $gte, $in, $nin, $and, $or, $not
+ - Bracket notation: ?filter[name][$like]=value
+ - Relationship traversal: ?filter[relationship][$exists]=true or ?filter[relationship][field]=value
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ WorkspacesPage
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.entity_store.list_workspaces()
+ """
+ _response = self._raw_client.list_workspaces(
+ page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ def create_workspace(
+ self,
+ *,
+ name: str,
+ wait_role_propagation: typing.Optional[bool] = None,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Workspace:
+ """
+ Create a new workspace.
+
+ The creator is automatically granted Admin role on the workspace.
+ By default, this endpoint waits for the Admin role to propagate before returning.
+ Use `wait_role_propagation=false` to skip waiting (useful for bulk operations).
+
+ Example:
+ ```
+ POST /apis/entities/v2/workspaces
+ {
+ "name": "ml-team",
+ "description": "Machine Learning Team workspace"
+ }
+ ```
+
+ Parameters
+ ----------
+ name : str
+ Workspace name (unique identifier). Name must start with a lowercase letter, be 2-63 characters, and contain only lowercase letters, digits, and hyphens (no consecutive hyphens, cannot end with a hyphen).
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for Admin role to propagate before returning (default: true). Set to false for bulk operations.
+
+ description : typing.Optional[str]
+ Optional description of the workspace
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Workspace
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.entity_store.create_workspace(
+ name="ml-team",
+ )
+ """
+ _response = self._raw_client.create_workspace(
+ name=name,
+ wait_role_propagation=wait_role_propagation,
+ description=description,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def get_workspace(self, name: str, *, request_options: typing.Optional[RequestOptions] = None) -> Workspace:
+ """
+ Get a specific workspace by ID.
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces/ml-team
+ ```
+
+ Parameters
+ ----------
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Workspace
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.entity_store.get_workspace(
+ name="name",
+ )
+ """
+ _response = self._raw_client.get_workspace(name, request_options=request_options)
+ return _response.data
+
+ def update_workspace(
+ self,
+ name: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Workspace:
+ """
+ Update a workspace's description.
+
+ Example:
+ ```
+ PUT /apis/entities/v2/workspaces/ml-team
+ {
+ "description": "Updated description for ML Team"
+ }
+ ```
+
+ Parameters
+ ----------
+ name : str
+
+ description : typing.Optional[str]
+ Updated description
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Workspace
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.entity_store.update_workspace(
+ name="name",
+ )
+ """
+ _response = self._raw_client.update_workspace(name, description=description, request_options=request_options)
+ return _response.data
+
+ def delete_workspace(self, name: str, *, request_options: typing.Optional[RequestOptions] = None) -> DeleteResponse:
+ """
+ Delete a workspace.
+
+ This marks the workspace for deletion and returns immediately. The workspace
+ will no longer be accessible via the API. An asynchronous cleanup controller
+ will handle deletion of all entities and external resources.
+
+ Role bindings are immediately deleted to revoke access.
+
+ Example:
+ ```
+ DELETE /apis/entities/v2/workspaces/ml-team
+ ```
+
+ Parameters
+ ----------
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ DeleteResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.entity_store.delete_workspace(
+ name="name",
+ )
+ """
+ _response = self._raw_client.delete_workspace(name, request_options=request_options)
+ return _response.data
+
+ def list_entities(
+ self,
+ workspace: str,
+ entity_type: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[str] = None,
+ filter: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> EntitiesPage:
+ """
+ List all entities of a specific type in the given workspace.
+
+ Use workspace="-" to list entities across all workspaces the principal has
+ access to.
+
+ Query Parameters:
+ - sort: Sort field
+ - page, page_size: Pagination
+ - filter: Advanced filters (JSON, text, or bracket notation)
+
+ Examples:
+ ```
+ GET /apis/entities/v2/workspaces/default/entities/customization_config?sort=-created_at
+ GET /apis/entities/v2/workspaces/-/entities/customization_config # Cross-workspace query
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ entity_type : str
+
+ page : typing.Optional[int]
+ Page number
+
+ page_size : typing.Optional[int]
+ Items per page
+
+ sort : typing.Optional[str]
+ Sort field
+
+ filter : typing.Optional[str]
+ Query filter expression. Supports text and JSON syntaxes:
+ - Text: name:"value" AND status>500 with operators : ~ > >= < <= IN NOT IN AND OR and negation prefix -
+ - Object (JSON): {"name":{"$like":"value"}} with operators $eq, $like, $lt, $lte, $gt, $gte, $in, $nin, $and, $or, $not
+ - Bracket notation: ?filter[name][$like]=value
+ - Relationship traversal: ?filter[relationship][$exists]=true or ?filter[relationship][field]=value
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ EntitiesPage
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.entity_store.list_entities(
+ workspace="workspace",
+ entity_type="entity_type",
+ sort="-created_at",
+ )
+ """
+ _response = self._raw_client.list_entities(
+ workspace,
+ entity_type,
+ page=page,
+ page_size=page_size,
+ sort=sort,
+ filter=filter,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def create_entity(
+ self,
+ workspace: str,
+ entity_type: str,
+ *,
+ data: typing.Dict[str, typing.Any],
+ name: typing.Optional[str] = OMIT,
+ parent: typing.Optional[str] = OMIT,
+ project: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Entity:
+ """
+ Create a new entity of the specified type in the given workspace.
+
+ If name is not provided, it will be auto-generated based on the entity type.
+
+ Example:
+ ```
+ POST /apis/entities/v2/workspaces/default/entities/customization_config
+ {
+ "name": "my-config",
+ "data": {
+ "target_id": "llama-2-7b",
+ "training_options": {"learning_rate": 0.01}
+ }
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ entity_type : str
+
+ data : typing.Dict[str, typing.Any]
+ Entity-specific data (schema is opaque to entity store, validated by client SDK)
+
+ name : typing.Optional[str]
+ Entity name (optional - auto-generated if not provided). Name must start with a lowercase letter, be 2-63 characters, and contain only lowercase letters, digits, and hyphens (no consecutive hyphens, cannot end with a hyphen).
+
+ parent : typing.Optional[str]
+ Parent entity ID for nested entities
+
+ project : typing.Optional[str]
+ The name of the project associated with this entity
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Entity
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.entity_store.create_entity(
+ workspace="workspace",
+ entity_type="entity_type",
+ data={"key": "value"},
+ )
+ """
+ _response = self._raw_client.create_entity(
+ workspace,
+ entity_type,
+ data=data,
+ name=name,
+ parent=parent,
+ project=project,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def get_entity_by_name(
+ self,
+ workspace: str,
+ entity_type: str,
+ name: str,
+ *,
+ parent: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Entity:
+ """
+ Get a specific entity by its workspace, type, and name.
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces/default/entities/customization_config/my-config
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ entity_type : str
+
+ name : str
+
+ parent : typing.Optional[str]
+ Parent entity ID for nested entities
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Entity
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.entity_store.get_entity_by_name(
+ workspace="workspace",
+ entity_type="entity_type",
+ name="name",
+ )
+ """
+ _response = self._raw_client.get_entity_by_name(
+ workspace, entity_type, name, parent=parent, request_options=request_options
+ )
+ return _response.data
+
+ def update_entity_by_name(
+ self,
+ workspace: str,
+ entity_type: str,
+ name: str,
+ *,
+ data: typing.Dict[str, typing.Any],
+ parent: typing.Optional[str] = None,
+ new_name: typing.Optional[str] = OMIT,
+ project: typing.Optional[str] = OMIT,
+ expected_db_version: typing.Optional[int] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Entity:
+ """
+ Update an entity by its name. Optionally change the entity's name.
+
+ Example:
+ ```
+ PUT /apis/entities/v2/workspaces/default/entities/customization_config/my-config
+ {
+ "data": {
+ "target_id": "llama-2-7b",
+ "training_options": {"learning_rate": 0.02}
+ }
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ entity_type : str
+
+ name : str
+
+ data : typing.Dict[str, typing.Any]
+ Updated entity-specific data
+
+ parent : typing.Optional[str]
+ Parent entity ID for nested entities
+
+ new_name : typing.Optional[str]
+ Updated entity name (optional). Name must start with a lowercase letter, be 2-63 characters, and contain only lowercase letters, digits, and hyphens (no consecutive hyphens, cannot end with a hyphen).
+
+ project : typing.Optional[str]
+ The name of the project associated with this entity
+
+ expected_db_version : typing.Optional[int]
+ Optional database version for optimistic locking. Update only succeeds if current version matches.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Entity
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.entity_store.update_entity_by_name(
+ workspace="workspace",
+ entity_type="entity_type",
+ name="name",
+ data={"key": "value"},
+ )
+ """
+ _response = self._raw_client.update_entity_by_name(
+ workspace,
+ entity_type,
+ name,
+ data=data,
+ parent=parent,
+ new_name=new_name,
+ project=project,
+ expected_db_version=expected_db_version,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def delete_entity_by_name(
+ self,
+ workspace: str,
+ entity_type: str,
+ name: str,
+ *,
+ parent: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> DeleteResponse:
+ """
+ Delete an entity by its name.
+
+ Example:
+ ```
+ DELETE /apis/entities/v2/workspaces/default/entities/customization_config/my-config
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ entity_type : str
+
+ name : str
+
+ parent : typing.Optional[str]
+ Parent entity ID for nested entities
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ DeleteResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.entity_store.delete_entity_by_name(
+ workspace="workspace",
+ entity_type="entity_type",
+ name="name",
+ )
+ """
+ _response = self._raw_client.delete_entity_by_name(
+ workspace, entity_type, name, parent=parent, request_options=request_options
+ )
+ return _response.data
+
+ def list_workspace_members(
+ self, workspace: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> WorkspaceMemberListResponse:
+ """
+ List all members of a workspace with their roles.
+
+ Returns a list of all principals with active role bindings in the workspace.
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces/ml-team/members
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ WorkspaceMemberListResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.entity_store.list_workspace_members(
+ workspace="workspace",
+ )
+ """
+ _response = self._raw_client.list_workspace_members(workspace, request_options=request_options)
+ return _response.data
+
+ def add_workspace_member(
+ self,
+ workspace: str,
+ *,
+ principal: str,
+ wait_role_propagation: typing.Optional[bool] = None,
+ roles: typing.Optional[typing.Sequence[str]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> WorkspaceMember:
+ """
+ Add a new member to the workspace with specified roles.
+
+ This creates role bindings for the specified principal with the given roles.
+ By default, this endpoint waits for the roles to propagate before returning.
+ Use `wait_role_propagation=false` to skip waiting (useful for bulk operations).
+
+ Example:
+ ```
+ POST /apis/entities/v2/workspaces/ml-team/members
+ {
+ "principal": "user@example.com",
+ "roles": ["Editor"]
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ principal : str
+ The principal identifier (email, user ID, or group ID)
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for roles to propagate before returning (default: true). Set to false for bulk operations.
+
+ roles : typing.Optional[typing.Sequence[str]]
+ List of roles to grant to the principal
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ WorkspaceMember
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.entity_store.add_workspace_member(
+ workspace="workspace",
+ principal="user@example.com",
+ )
+ """
+ _response = self._raw_client.add_workspace_member(
+ workspace,
+ principal=principal,
+ wait_role_propagation=wait_role_propagation,
+ roles=roles,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def update_workspace_member(
+ self,
+ workspace: str,
+ principal_id: str,
+ *,
+ roles: typing.Sequence[str],
+ wait_role_propagation: typing.Optional[bool] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> WorkspaceMember:
+ """
+ Update the roles for a workspace member.
+
+ This will revoke existing roles not in the new list and add new roles.
+ By default, this endpoint waits for the roles to propagate before returning.
+ Use `wait_role_propagation=false` to skip waiting (useful for bulk operations).
+
+ Example:
+ ```
+ PUT /apis/entities/v2/workspaces/ml-team/members/user@example.com
+ {
+ "roles": ["Viewer", "Editor"]
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ principal_id : str
+
+ roles : typing.Sequence[str]
+ Updated list of roles for the principal
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for roles to propagate before returning (default: true). Set to false for bulk operations.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ WorkspaceMember
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.entity_store.update_workspace_member(
+ workspace="workspace",
+ principal_id="principal_id",
+ roles=["Viewer"],
+ )
+ """
+ _response = self._raw_client.update_workspace_member(
+ workspace,
+ principal_id,
+ roles=roles,
+ wait_role_propagation=wait_role_propagation,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def remove_workspace_member(
+ self,
+ workspace: str,
+ principal_id: str,
+ *,
+ wait_role_propagation: typing.Optional[bool] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> DeleteResponse:
+ """
+ Remove a member from the workspace by revoking all their roles.
+
+ This revokes all active role bindings for the principal in the workspace.
+ By default, this endpoint waits for all roles to be revoked before returning.
+ Use `wait_role_propagation=false` to skip waiting (useful for bulk operations).
+
+ Example:
+ ```
+ DELETE /apis/entities/v2/workspaces/ml-team/members/user@example.com
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ principal_id : str
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for roles to propagate before returning (default: true). Set to false for bulk operations.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ DeleteResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.entity_store.remove_workspace_member(
+ workspace="workspace",
+ principal_id="principal_id",
+ )
+ """
+ _response = self._raw_client.remove_workspace_member(
+ workspace, principal_id, wait_role_propagation=wait_role_propagation, request_options=request_options
+ )
+ return _response.data
+
+ def list_projects(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[ProjectSortField] = None,
+ filter: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ProjectsPage:
+ """
+ List all projects in a workspace with pagination.
+
+ Query Parameters:
+ - page, page_size: Pagination
+ - sort: Sort field
+ - filter: Advanced filters
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces/default/projects?sort=-created_at&page=1&page_size=10
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number
+
+ page_size : typing.Optional[int]
+ Items per page
+
+ sort : typing.Optional[ProjectSortField]
+ Sort field
+
+ filter : typing.Optional[str]
+ Query filter expression. Supports text and JSON syntaxes:
+ - Text: name:"value" AND status>500 with operators : ~ > >= < <= IN NOT IN AND OR and negation prefix -
+ - Object (JSON): {"name":{"$like":"value"}} with operators $eq, $like, $lt, $lte, $gt, $gte, $in, $nin, $and, $or, $not
+ - Bracket notation: ?filter[name][$like]=value
+ - Relationship traversal: ?filter[relationship][$exists]=true or ?filter[relationship][field]=value
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ProjectsPage
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.entity_store.list_projects(
+ workspace="workspace",
+ )
+ """
+ _response = self._raw_client.list_projects(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ def create_project(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Project:
+ """
+ Create a new project in the given workspace.
+
+ Example:
+ ```
+ POST /apis/entities/v2/workspaces/default/projects
+ {
+ "name": "ml-project",
+ "description": "Machine Learning project"
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Project name (unique within workspace). Name must start with a lowercase letter, be 2-63 characters, and contain only lowercase letters, digits, and hyphens (no consecutive hyphens, cannot end with a hyphen).
+
+ description : typing.Optional[str]
+ Optional description of the project
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Project
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.entity_store.create_project(
+ workspace="workspace",
+ name="ml-project",
+ )
+ """
+ _response = self._raw_client.create_project(
+ workspace, name=name, description=description, request_options=request_options
+ )
+ return _response.data
+
+ def get_project(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> Project:
+ """
+ Get a specific project by its workspace and name.
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces/default/projects/ml-project
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Project
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.entity_store.get_project(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.get_project(workspace, name, request_options=request_options)
+ return _response.data
+
+ def update_project(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Project:
+ """
+ Update a project's description.
+
+ Example:
+ ```
+ PUT /apis/entities/v2/workspaces/default/projects/ml-project
+ {
+ "description": "Updated description for ML project"
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ description : typing.Optional[str]
+ Updated description
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Project
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.entity_store.update_project(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.update_project(
+ workspace, name, description=description, request_options=request_options
+ )
+ return _response.data
+
+ def delete_project(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> DeleteResponse:
+ """
+ Delete a project.
+
+ Example:
+ ```
+ DELETE /apis/entities/v2/workspaces/default/projects/ml-project
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ DeleteResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.entity_store.delete_project(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.delete_project(workspace, name, request_options=request_options)
+ return _response.data
+
+
+class AsyncEntityStoreClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._raw_client = AsyncRawEntityStoreClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> AsyncRawEntityStoreClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ AsyncRawEntityStoreClient
+ """
+ return self._raw_client
+
+ async def get_entity_by_id(self, id: str, *, request_options: typing.Optional[RequestOptions] = None) -> Entity:
+ """
+ Get a specific entity by its unique identifier.
+ This endpoint is primarily for debugging and internal use.
+
+ Example:
+ ```
+ GET /apis/entities/v2/entities/customization-config-5Q2LoF8z8M9JZxZsHwJKNn
+ ```
+
+ Parameters
+ ----------
+ id : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Entity
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.entity_store.get_entity_by_id(
+ id="id",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_entity_by_id(id, request_options=request_options)
+ return _response.data
+
+ async def list_workspaces(
+ self,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[GenericSortField] = None,
+ filter: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> WorkspacesPage:
+ """
+ List all workspaces with pagination.
+
+ When authentication is enabled, only workspaces the principal has access to
+ are returned. Service principals and platform admins have access to all workspaces.
+
+ Query Parameters:
+ - page, page_size: Pagination
+ - sort: Sort field
+ - filter: Advanced filters (JSON, text, or bracket notation)
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces?sort=-created_at&page=1&page_size=10
+ ```
+
+ Parameters
+ ----------
+ page : typing.Optional[int]
+ Page number
+
+ page_size : typing.Optional[int]
+ Items per page
+
+ sort : typing.Optional[GenericSortField]
+ Sort field
+
+ filter : typing.Optional[str]
+ Query filter expression. Supports text and JSON syntaxes:
+ - Text: name:"value" AND status>500 with operators : ~ > >= < <= IN NOT IN AND OR and negation prefix -
+ - Object (JSON): {"name":{"$like":"value"}} with operators $eq, $like, $lt, $lte, $gt, $gte, $in, $nin, $and, $or, $not
+ - Bracket notation: ?filter[name][$like]=value
+ - Relationship traversal: ?filter[relationship][$exists]=true or ?filter[relationship][field]=value
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ WorkspacesPage
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.entity_store.list_workspaces()
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_workspaces(
+ page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ async def create_workspace(
+ self,
+ *,
+ name: str,
+ wait_role_propagation: typing.Optional[bool] = None,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Workspace:
+ """
+ Create a new workspace.
+
+ The creator is automatically granted Admin role on the workspace.
+ By default, this endpoint waits for the Admin role to propagate before returning.
+ Use `wait_role_propagation=false` to skip waiting (useful for bulk operations).
+
+ Example:
+ ```
+ POST /apis/entities/v2/workspaces
+ {
+ "name": "ml-team",
+ "description": "Machine Learning Team workspace"
+ }
+ ```
+
+ Parameters
+ ----------
+ name : str
+ Workspace name (unique identifier). Name must start with a lowercase letter, be 2-63 characters, and contain only lowercase letters, digits, and hyphens (no consecutive hyphens, cannot end with a hyphen).
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for Admin role to propagate before returning (default: true). Set to false for bulk operations.
+
+ description : typing.Optional[str]
+ Optional description of the workspace
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Workspace
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.entity_store.create_workspace(
+ name="ml-team",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.create_workspace(
+ name=name,
+ wait_role_propagation=wait_role_propagation,
+ description=description,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def get_workspace(self, name: str, *, request_options: typing.Optional[RequestOptions] = None) -> Workspace:
+ """
+ Get a specific workspace by ID.
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces/ml-team
+ ```
+
+ Parameters
+ ----------
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Workspace
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.entity_store.get_workspace(
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_workspace(name, request_options=request_options)
+ return _response.data
+
+ async def update_workspace(
+ self,
+ name: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Workspace:
+ """
+ Update a workspace's description.
+
+ Example:
+ ```
+ PUT /apis/entities/v2/workspaces/ml-team
+ {
+ "description": "Updated description for ML Team"
+ }
+ ```
+
+ Parameters
+ ----------
+ name : str
+
+ description : typing.Optional[str]
+ Updated description
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Workspace
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.entity_store.update_workspace(
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.update_workspace(
+ name, description=description, request_options=request_options
+ )
+ return _response.data
+
+ async def delete_workspace(
+ self, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> DeleteResponse:
+ """
+ Delete a workspace.
+
+ This marks the workspace for deletion and returns immediately. The workspace
+ will no longer be accessible via the API. An asynchronous cleanup controller
+ will handle deletion of all entities and external resources.
+
+ Role bindings are immediately deleted to revoke access.
+
+ Example:
+ ```
+ DELETE /apis/entities/v2/workspaces/ml-team
+ ```
+
+ Parameters
+ ----------
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ DeleteResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.entity_store.delete_workspace(
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.delete_workspace(name, request_options=request_options)
+ return _response.data
+
+ async def list_entities(
+ self,
+ workspace: str,
+ entity_type: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[str] = None,
+ filter: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> EntitiesPage:
+ """
+ List all entities of a specific type in the given workspace.
+
+ Use workspace="-" to list entities across all workspaces the principal has
+ access to.
+
+ Query Parameters:
+ - sort: Sort field
+ - page, page_size: Pagination
+ - filter: Advanced filters (JSON, text, or bracket notation)
+
+ Examples:
+ ```
+ GET /apis/entities/v2/workspaces/default/entities/customization_config?sort=-created_at
+ GET /apis/entities/v2/workspaces/-/entities/customization_config # Cross-workspace query
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ entity_type : str
+
+ page : typing.Optional[int]
+ Page number
+
+ page_size : typing.Optional[int]
+ Items per page
+
+ sort : typing.Optional[str]
+ Sort field
+
+ filter : typing.Optional[str]
+ Query filter expression. Supports text and JSON syntaxes:
+ - Text: name:"value" AND status>500 with operators : ~ > >= < <= IN NOT IN AND OR and negation prefix -
+ - Object (JSON): {"name":{"$like":"value"}} with operators $eq, $like, $lt, $lte, $gt, $gte, $in, $nin, $and, $or, $not
+ - Bracket notation: ?filter[name][$like]=value
+ - Relationship traversal: ?filter[relationship][$exists]=true or ?filter[relationship][field]=value
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ EntitiesPage
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.entity_store.list_entities(
+ workspace="workspace",
+ entity_type="entity_type",
+ sort="-created_at",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_entities(
+ workspace,
+ entity_type,
+ page=page,
+ page_size=page_size,
+ sort=sort,
+ filter=filter,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def create_entity(
+ self,
+ workspace: str,
+ entity_type: str,
+ *,
+ data: typing.Dict[str, typing.Any],
+ name: typing.Optional[str] = OMIT,
+ parent: typing.Optional[str] = OMIT,
+ project: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Entity:
+ """
+ Create a new entity of the specified type in the given workspace.
+
+ If name is not provided, it will be auto-generated based on the entity type.
+
+ Example:
+ ```
+ POST /apis/entities/v2/workspaces/default/entities/customization_config
+ {
+ "name": "my-config",
+ "data": {
+ "target_id": "llama-2-7b",
+ "training_options": {"learning_rate": 0.01}
+ }
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ entity_type : str
+
+ data : typing.Dict[str, typing.Any]
+ Entity-specific data (schema is opaque to entity store, validated by client SDK)
+
+ name : typing.Optional[str]
+ Entity name (optional - auto-generated if not provided). Name must start with a lowercase letter, be 2-63 characters, and contain only lowercase letters, digits, and hyphens (no consecutive hyphens, cannot end with a hyphen).
+
+ parent : typing.Optional[str]
+ Parent entity ID for nested entities
+
+ project : typing.Optional[str]
+ The name of the project associated with this entity
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Entity
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.entity_store.create_entity(
+ workspace="workspace",
+ entity_type="entity_type",
+ data={"key": "value"},
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.create_entity(
+ workspace,
+ entity_type,
+ data=data,
+ name=name,
+ parent=parent,
+ project=project,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def get_entity_by_name(
+ self,
+ workspace: str,
+ entity_type: str,
+ name: str,
+ *,
+ parent: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Entity:
+ """
+ Get a specific entity by its workspace, type, and name.
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces/default/entities/customization_config/my-config
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ entity_type : str
+
+ name : str
+
+ parent : typing.Optional[str]
+ Parent entity ID for nested entities
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Entity
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.entity_store.get_entity_by_name(
+ workspace="workspace",
+ entity_type="entity_type",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_entity_by_name(
+ workspace, entity_type, name, parent=parent, request_options=request_options
+ )
+ return _response.data
+
+ async def update_entity_by_name(
+ self,
+ workspace: str,
+ entity_type: str,
+ name: str,
+ *,
+ data: typing.Dict[str, typing.Any],
+ parent: typing.Optional[str] = None,
+ new_name: typing.Optional[str] = OMIT,
+ project: typing.Optional[str] = OMIT,
+ expected_db_version: typing.Optional[int] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Entity:
+ """
+ Update an entity by its name. Optionally change the entity's name.
+
+ Example:
+ ```
+ PUT /apis/entities/v2/workspaces/default/entities/customization_config/my-config
+ {
+ "data": {
+ "target_id": "llama-2-7b",
+ "training_options": {"learning_rate": 0.02}
+ }
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ entity_type : str
+
+ name : str
+
+ data : typing.Dict[str, typing.Any]
+ Updated entity-specific data
+
+ parent : typing.Optional[str]
+ Parent entity ID for nested entities
+
+ new_name : typing.Optional[str]
+ Updated entity name (optional). Name must start with a lowercase letter, be 2-63 characters, and contain only lowercase letters, digits, and hyphens (no consecutive hyphens, cannot end with a hyphen).
+
+ project : typing.Optional[str]
+ The name of the project associated with this entity
+
+ expected_db_version : typing.Optional[int]
+ Optional database version for optimistic locking. Update only succeeds if current version matches.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Entity
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.entity_store.update_entity_by_name(
+ workspace="workspace",
+ entity_type="entity_type",
+ name="name",
+ data={"key": "value"},
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.update_entity_by_name(
+ workspace,
+ entity_type,
+ name,
+ data=data,
+ parent=parent,
+ new_name=new_name,
+ project=project,
+ expected_db_version=expected_db_version,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def delete_entity_by_name(
+ self,
+ workspace: str,
+ entity_type: str,
+ name: str,
+ *,
+ parent: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> DeleteResponse:
+ """
+ Delete an entity by its name.
+
+ Example:
+ ```
+ DELETE /apis/entities/v2/workspaces/default/entities/customization_config/my-config
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ entity_type : str
+
+ name : str
+
+ parent : typing.Optional[str]
+ Parent entity ID for nested entities
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ DeleteResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.entity_store.delete_entity_by_name(
+ workspace="workspace",
+ entity_type="entity_type",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.delete_entity_by_name(
+ workspace, entity_type, name, parent=parent, request_options=request_options
+ )
+ return _response.data
+
+ async def list_workspace_members(
+ self, workspace: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> WorkspaceMemberListResponse:
+ """
+ List all members of a workspace with their roles.
+
+ Returns a list of all principals with active role bindings in the workspace.
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces/ml-team/members
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ WorkspaceMemberListResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.entity_store.list_workspace_members(
+ workspace="workspace",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_workspace_members(workspace, request_options=request_options)
+ return _response.data
+
+ async def add_workspace_member(
+ self,
+ workspace: str,
+ *,
+ principal: str,
+ wait_role_propagation: typing.Optional[bool] = None,
+ roles: typing.Optional[typing.Sequence[str]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> WorkspaceMember:
+ """
+ Add a new member to the workspace with specified roles.
+
+ This creates role bindings for the specified principal with the given roles.
+ By default, this endpoint waits for the roles to propagate before returning.
+ Use `wait_role_propagation=false` to skip waiting (useful for bulk operations).
+
+ Example:
+ ```
+ POST /apis/entities/v2/workspaces/ml-team/members
+ {
+ "principal": "user@example.com",
+ "roles": ["Editor"]
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ principal : str
+ The principal identifier (email, user ID, or group ID)
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for roles to propagate before returning (default: true). Set to false for bulk operations.
+
+ roles : typing.Optional[typing.Sequence[str]]
+ List of roles to grant to the principal
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ WorkspaceMember
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.entity_store.add_workspace_member(
+ workspace="workspace",
+ principal="user@example.com",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.add_workspace_member(
+ workspace,
+ principal=principal,
+ wait_role_propagation=wait_role_propagation,
+ roles=roles,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def update_workspace_member(
+ self,
+ workspace: str,
+ principal_id: str,
+ *,
+ roles: typing.Sequence[str],
+ wait_role_propagation: typing.Optional[bool] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> WorkspaceMember:
+ """
+ Update the roles for a workspace member.
+
+ This will revoke existing roles not in the new list and add new roles.
+ By default, this endpoint waits for the roles to propagate before returning.
+ Use `wait_role_propagation=false` to skip waiting (useful for bulk operations).
+
+ Example:
+ ```
+ PUT /apis/entities/v2/workspaces/ml-team/members/user@example.com
+ {
+ "roles": ["Viewer", "Editor"]
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ principal_id : str
+
+ roles : typing.Sequence[str]
+ Updated list of roles for the principal
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for roles to propagate before returning (default: true). Set to false for bulk operations.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ WorkspaceMember
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.entity_store.update_workspace_member(
+ workspace="workspace",
+ principal_id="principal_id",
+ roles=["Viewer"],
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.update_workspace_member(
+ workspace,
+ principal_id,
+ roles=roles,
+ wait_role_propagation=wait_role_propagation,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def remove_workspace_member(
+ self,
+ workspace: str,
+ principal_id: str,
+ *,
+ wait_role_propagation: typing.Optional[bool] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> DeleteResponse:
+ """
+ Remove a member from the workspace by revoking all their roles.
+
+ This revokes all active role bindings for the principal in the workspace.
+ By default, this endpoint waits for all roles to be revoked before returning.
+ Use `wait_role_propagation=false` to skip waiting (useful for bulk operations).
+
+ Example:
+ ```
+ DELETE /apis/entities/v2/workspaces/ml-team/members/user@example.com
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ principal_id : str
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for roles to propagate before returning (default: true). Set to false for bulk operations.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ DeleteResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.entity_store.remove_workspace_member(
+ workspace="workspace",
+ principal_id="principal_id",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.remove_workspace_member(
+ workspace, principal_id, wait_role_propagation=wait_role_propagation, request_options=request_options
+ )
+ return _response.data
+
+ async def list_projects(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[ProjectSortField] = None,
+ filter: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ProjectsPage:
+ """
+ List all projects in a workspace with pagination.
+
+ Query Parameters:
+ - page, page_size: Pagination
+ - sort: Sort field
+ - filter: Advanced filters
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces/default/projects?sort=-created_at&page=1&page_size=10
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number
+
+ page_size : typing.Optional[int]
+ Items per page
+
+ sort : typing.Optional[ProjectSortField]
+ Sort field
+
+ filter : typing.Optional[str]
+ Query filter expression. Supports text and JSON syntaxes:
+ - Text: name:"value" AND status>500 with operators : ~ > >= < <= IN NOT IN AND OR and negation prefix -
+ - Object (JSON): {"name":{"$like":"value"}} with operators $eq, $like, $lt, $lte, $gt, $gte, $in, $nin, $and, $or, $not
+ - Bracket notation: ?filter[name][$like]=value
+ - Relationship traversal: ?filter[relationship][$exists]=true or ?filter[relationship][field]=value
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ProjectsPage
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.entity_store.list_projects(
+ workspace="workspace",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_projects(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ async def create_project(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Project:
+ """
+ Create a new project in the given workspace.
+
+ Example:
+ ```
+ POST /apis/entities/v2/workspaces/default/projects
+ {
+ "name": "ml-project",
+ "description": "Machine Learning project"
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Project name (unique within workspace). Name must start with a lowercase letter, be 2-63 characters, and contain only lowercase letters, digits, and hyphens (no consecutive hyphens, cannot end with a hyphen).
+
+ description : typing.Optional[str]
+ Optional description of the project
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Project
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.entity_store.create_project(
+ workspace="workspace",
+ name="ml-project",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.create_project(
+ workspace, name=name, description=description, request_options=request_options
+ )
+ return _response.data
+
+ async def get_project(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> Project:
+ """
+ Get a specific project by its workspace and name.
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces/default/projects/ml-project
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Project
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.entity_store.get_project(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_project(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def update_project(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Project:
+ """
+ Update a project's description.
+
+ Example:
+ ```
+ PUT /apis/entities/v2/workspaces/default/projects/ml-project
+ {
+ "description": "Updated description for ML project"
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ description : typing.Optional[str]
+ Updated description
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Project
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.entity_store.update_project(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.update_project(
+ workspace, name, description=description, request_options=request_options
+ )
+ return _response.data
+
+ async def delete_project(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> DeleteResponse:
+ """
+ Delete a project.
+
+ Example:
+ ```
+ DELETE /apis/entities/v2/workspaces/default/projects/ml-project
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ DeleteResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.entity_store.delete_project(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.delete_project(workspace, name, request_options=request_options)
+ return _response.data
diff --git a/sdks/python/entity_store/raw_client.py b/sdks/python/entity_store/raw_client.py
new file mode 100644
index 0000000000..6d37a1277e
--- /dev/null
+++ b/sdks/python/entity_store/raw_client.py
@@ -0,0 +1,3190 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+from json.decoder import JSONDecodeError
+
+from ..core.api_error import ApiError
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.http_response import AsyncHttpResponse, HttpResponse
+from ..core.jsonable_encoder import encode_path_param
+from ..core.parse_error import ParsingError
+from ..core.pydantic_utilities import parse_obj_as
+from ..core.request_options import RequestOptions
+from ..errors.unprocessable_entity_error import UnprocessableEntityError
+from ..types.delete_response import DeleteResponse
+from ..types.entities_page import EntitiesPage
+from ..types.entity import Entity
+from ..types.generic_sort_field import GenericSortField
+from ..types.project import Project
+from ..types.project_sort_field import ProjectSortField
+from ..types.projects_page import ProjectsPage
+from ..types.workspace import Workspace
+from ..types.workspace_member import WorkspaceMember
+from ..types.workspace_member_list_response import WorkspaceMemberListResponse
+from ..types.workspaces_page import WorkspacesPage
+from pydantic import ValidationError
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class RawEntityStoreClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ def get_entity_by_id(
+ self, id: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[Entity]:
+ """
+ Get a specific entity by its unique identifier.
+ This endpoint is primarily for debugging and internal use.
+
+ Example:
+ ```
+ GET /apis/entities/v2/entities/customization-config-5Q2LoF8z8M9JZxZsHwJKNn
+ ```
+
+ Parameters
+ ----------
+ id : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[Entity]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/entities/{encode_path_param(id)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Entity,
+ parse_obj_as(
+ type_=Entity, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def list_workspaces(
+ self,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[GenericSortField] = None,
+ filter: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[WorkspacesPage]:
+ """
+ List all workspaces with pagination.
+
+ When authentication is enabled, only workspaces the principal has access to
+ are returned. Service principals and platform admins have access to all workspaces.
+
+ Query Parameters:
+ - page, page_size: Pagination
+ - sort: Sort field
+ - filter: Advanced filters (JSON, text, or bracket notation)
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces?sort=-created_at&page=1&page_size=10
+ ```
+
+ Parameters
+ ----------
+ page : typing.Optional[int]
+ Page number
+
+ page_size : typing.Optional[int]
+ Items per page
+
+ sort : typing.Optional[GenericSortField]
+ Sort field
+
+ filter : typing.Optional[str]
+ Query filter expression. Supports text and JSON syntaxes:
+ - Text: name:"value" AND status>500 with operators : ~ > >= < <= IN NOT IN AND OR and negation prefix -
+ - Object (JSON): {"name":{"$like":"value"}} with operators $eq, $like, $lt, $lte, $gt, $gte, $in, $nin, $and, $or, $not
+ - Bracket notation: ?filter[name][$like]=value
+ - Relationship traversal: ?filter[relationship][$exists]=true or ?filter[relationship][field]=value
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[WorkspacesPage]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ "apis/entities/v2/workspaces",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": filter,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ WorkspacesPage,
+ parse_obj_as(
+ type_=WorkspacesPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def create_workspace(
+ self,
+ *,
+ name: str,
+ wait_role_propagation: typing.Optional[bool] = None,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[Workspace]:
+ """
+ Create a new workspace.
+
+ The creator is automatically granted Admin role on the workspace.
+ By default, this endpoint waits for the Admin role to propagate before returning.
+ Use `wait_role_propagation=false` to skip waiting (useful for bulk operations).
+
+ Example:
+ ```
+ POST /apis/entities/v2/workspaces
+ {
+ "name": "ml-team",
+ "description": "Machine Learning Team workspace"
+ }
+ ```
+
+ Parameters
+ ----------
+ name : str
+ Workspace name (unique identifier). Name must start with a lowercase letter, be 2-63 characters, and contain only lowercase letters, digits, and hyphens (no consecutive hyphens, cannot end with a hyphen).
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for Admin role to propagate before returning (default: true). Set to false for bulk operations.
+
+ description : typing.Optional[str]
+ Optional description of the workspace
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[Workspace]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ "apis/entities/v2/workspaces",
+ method="POST",
+ params={
+ "wait_role_propagation": wait_role_propagation,
+ },
+ json={
+ "name": name,
+ "description": description,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Workspace,
+ parse_obj_as(
+ type_=Workspace, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_workspace(
+ self, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[Workspace]:
+ """
+ Get a specific workspace by ID.
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces/ml-team
+ ```
+
+ Parameters
+ ----------
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[Workspace]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Workspace,
+ parse_obj_as(
+ type_=Workspace, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def update_workspace(
+ self,
+ name: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[Workspace]:
+ """
+ Update a workspace's description.
+
+ Example:
+ ```
+ PUT /apis/entities/v2/workspaces/ml-team
+ {
+ "description": "Updated description for ML Team"
+ }
+ ```
+
+ Parameters
+ ----------
+ name : str
+
+ description : typing.Optional[str]
+ Updated description
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[Workspace]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(name)}",
+ method="PUT",
+ json={
+ "description": description,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Workspace,
+ parse_obj_as(
+ type_=Workspace, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def delete_workspace(
+ self, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[DeleteResponse]:
+ """
+ Delete a workspace.
+
+ This marks the workspace for deletion and returns immediately. The workspace
+ will no longer be accessible via the API. An asynchronous cleanup controller
+ will handle deletion of all entities and external resources.
+
+ Role bindings are immediately deleted to revoke access.
+
+ Example:
+ ```
+ DELETE /apis/entities/v2/workspaces/ml-team
+ ```
+
+ Parameters
+ ----------
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[DeleteResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ DeleteResponse,
+ parse_obj_as(
+ type_=DeleteResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def list_entities(
+ self,
+ workspace: str,
+ entity_type: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[str] = None,
+ filter: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[EntitiesPage]:
+ """
+ List all entities of a specific type in the given workspace.
+
+ Use workspace="-" to list entities across all workspaces the principal has
+ access to.
+
+ Query Parameters:
+ - sort: Sort field
+ - page, page_size: Pagination
+ - filter: Advanced filters (JSON, text, or bracket notation)
+
+ Examples:
+ ```
+ GET /apis/entities/v2/workspaces/default/entities/customization_config?sort=-created_at
+ GET /apis/entities/v2/workspaces/-/entities/customization_config # Cross-workspace query
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ entity_type : str
+
+ page : typing.Optional[int]
+ Page number
+
+ page_size : typing.Optional[int]
+ Items per page
+
+ sort : typing.Optional[str]
+ Sort field
+
+ filter : typing.Optional[str]
+ Query filter expression. Supports text and JSON syntaxes:
+ - Text: name:"value" AND status>500 with operators : ~ > >= < <= IN NOT IN AND OR and negation prefix -
+ - Object (JSON): {"name":{"$like":"value"}} with operators $eq, $like, $lt, $lte, $gt, $gte, $in, $nin, $and, $or, $not
+ - Bracket notation: ?filter[name][$like]=value
+ - Relationship traversal: ?filter[relationship][$exists]=true or ?filter[relationship][field]=value
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[EntitiesPage]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/entities/{encode_path_param(entity_type)}",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": filter,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ EntitiesPage,
+ parse_obj_as(
+ type_=EntitiesPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def create_entity(
+ self,
+ workspace: str,
+ entity_type: str,
+ *,
+ data: typing.Dict[str, typing.Any],
+ name: typing.Optional[str] = OMIT,
+ parent: typing.Optional[str] = OMIT,
+ project: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[Entity]:
+ """
+ Create a new entity of the specified type in the given workspace.
+
+ If name is not provided, it will be auto-generated based on the entity type.
+
+ Example:
+ ```
+ POST /apis/entities/v2/workspaces/default/entities/customization_config
+ {
+ "name": "my-config",
+ "data": {
+ "target_id": "llama-2-7b",
+ "training_options": {"learning_rate": 0.01}
+ }
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ entity_type : str
+
+ data : typing.Dict[str, typing.Any]
+ Entity-specific data (schema is opaque to entity store, validated by client SDK)
+
+ name : typing.Optional[str]
+ Entity name (optional - auto-generated if not provided). Name must start with a lowercase letter, be 2-63 characters, and contain only lowercase letters, digits, and hyphens (no consecutive hyphens, cannot end with a hyphen).
+
+ parent : typing.Optional[str]
+ Parent entity ID for nested entities
+
+ project : typing.Optional[str]
+ The name of the project associated with this entity
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[Entity]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/entities/{encode_path_param(entity_type)}",
+ method="POST",
+ json={
+ "name": name,
+ "parent": parent,
+ "project": project,
+ "data": data,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Entity,
+ parse_obj_as(
+ type_=Entity, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_entity_by_name(
+ self,
+ workspace: str,
+ entity_type: str,
+ name: str,
+ *,
+ parent: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[Entity]:
+ """
+ Get a specific entity by its workspace, type, and name.
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces/default/entities/customization_config/my-config
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ entity_type : str
+
+ name : str
+
+ parent : typing.Optional[str]
+ Parent entity ID for nested entities
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[Entity]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/entities/{encode_path_param(entity_type)}/{encode_path_param(name)}",
+ method="GET",
+ params={
+ "parent": parent,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Entity,
+ parse_obj_as(
+ type_=Entity, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def update_entity_by_name(
+ self,
+ workspace: str,
+ entity_type: str,
+ name: str,
+ *,
+ data: typing.Dict[str, typing.Any],
+ parent: typing.Optional[str] = None,
+ new_name: typing.Optional[str] = OMIT,
+ project: typing.Optional[str] = OMIT,
+ expected_db_version: typing.Optional[int] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[Entity]:
+ """
+ Update an entity by its name. Optionally change the entity's name.
+
+ Example:
+ ```
+ PUT /apis/entities/v2/workspaces/default/entities/customization_config/my-config
+ {
+ "data": {
+ "target_id": "llama-2-7b",
+ "training_options": {"learning_rate": 0.02}
+ }
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ entity_type : str
+
+ name : str
+
+ data : typing.Dict[str, typing.Any]
+ Updated entity-specific data
+
+ parent : typing.Optional[str]
+ Parent entity ID for nested entities
+
+ new_name : typing.Optional[str]
+ Updated entity name (optional). Name must start with a lowercase letter, be 2-63 characters, and contain only lowercase letters, digits, and hyphens (no consecutive hyphens, cannot end with a hyphen).
+
+ project : typing.Optional[str]
+ The name of the project associated with this entity
+
+ expected_db_version : typing.Optional[int]
+ Optional database version for optimistic locking. Update only succeeds if current version matches.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[Entity]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/entities/{encode_path_param(entity_type)}/{encode_path_param(name)}",
+ method="PUT",
+ params={
+ "parent": parent,
+ },
+ json={
+ "new_name": new_name,
+ "project": project,
+ "data": data,
+ "expected_db_version": expected_db_version,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Entity,
+ parse_obj_as(
+ type_=Entity, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def delete_entity_by_name(
+ self,
+ workspace: str,
+ entity_type: str,
+ name: str,
+ *,
+ parent: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[DeleteResponse]:
+ """
+ Delete an entity by its name.
+
+ Example:
+ ```
+ DELETE /apis/entities/v2/workspaces/default/entities/customization_config/my-config
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ entity_type : str
+
+ name : str
+
+ parent : typing.Optional[str]
+ Parent entity ID for nested entities
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[DeleteResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/entities/{encode_path_param(entity_type)}/{encode_path_param(name)}",
+ method="DELETE",
+ params={
+ "parent": parent,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ DeleteResponse,
+ parse_obj_as(
+ type_=DeleteResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def list_workspace_members(
+ self, workspace: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[WorkspaceMemberListResponse]:
+ """
+ List all members of a workspace with their roles.
+
+ Returns a list of all principals with active role bindings in the workspace.
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces/ml-team/members
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[WorkspaceMemberListResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/members",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ WorkspaceMemberListResponse,
+ parse_obj_as(
+ type_=WorkspaceMemberListResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def add_workspace_member(
+ self,
+ workspace: str,
+ *,
+ principal: str,
+ wait_role_propagation: typing.Optional[bool] = None,
+ roles: typing.Optional[typing.Sequence[str]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[WorkspaceMember]:
+ """
+ Add a new member to the workspace with specified roles.
+
+ This creates role bindings for the specified principal with the given roles.
+ By default, this endpoint waits for the roles to propagate before returning.
+ Use `wait_role_propagation=false` to skip waiting (useful for bulk operations).
+
+ Example:
+ ```
+ POST /apis/entities/v2/workspaces/ml-team/members
+ {
+ "principal": "user@example.com",
+ "roles": ["Editor"]
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ principal : str
+ The principal identifier (email, user ID, or group ID)
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for roles to propagate before returning (default: true). Set to false for bulk operations.
+
+ roles : typing.Optional[typing.Sequence[str]]
+ List of roles to grant to the principal
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[WorkspaceMember]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/members",
+ method="POST",
+ params={
+ "wait_role_propagation": wait_role_propagation,
+ },
+ json={
+ "principal": principal,
+ "roles": roles,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ WorkspaceMember,
+ parse_obj_as(
+ type_=WorkspaceMember, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def update_workspace_member(
+ self,
+ workspace: str,
+ principal_id: str,
+ *,
+ roles: typing.Sequence[str],
+ wait_role_propagation: typing.Optional[bool] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[WorkspaceMember]:
+ """
+ Update the roles for a workspace member.
+
+ This will revoke existing roles not in the new list and add new roles.
+ By default, this endpoint waits for the roles to propagate before returning.
+ Use `wait_role_propagation=false` to skip waiting (useful for bulk operations).
+
+ Example:
+ ```
+ PUT /apis/entities/v2/workspaces/ml-team/members/user@example.com
+ {
+ "roles": ["Viewer", "Editor"]
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ principal_id : str
+
+ roles : typing.Sequence[str]
+ Updated list of roles for the principal
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for roles to propagate before returning (default: true). Set to false for bulk operations.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[WorkspaceMember]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/members/{encode_path_param(principal_id)}",
+ method="PUT",
+ params={
+ "wait_role_propagation": wait_role_propagation,
+ },
+ json={
+ "roles": roles,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ WorkspaceMember,
+ parse_obj_as(
+ type_=WorkspaceMember, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def remove_workspace_member(
+ self,
+ workspace: str,
+ principal_id: str,
+ *,
+ wait_role_propagation: typing.Optional[bool] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[DeleteResponse]:
+ """
+ Remove a member from the workspace by revoking all their roles.
+
+ This revokes all active role bindings for the principal in the workspace.
+ By default, this endpoint waits for all roles to be revoked before returning.
+ Use `wait_role_propagation=false` to skip waiting (useful for bulk operations).
+
+ Example:
+ ```
+ DELETE /apis/entities/v2/workspaces/ml-team/members/user@example.com
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ principal_id : str
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for roles to propagate before returning (default: true). Set to false for bulk operations.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[DeleteResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/members/{encode_path_param(principal_id)}",
+ method="DELETE",
+ params={
+ "wait_role_propagation": wait_role_propagation,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ DeleteResponse,
+ parse_obj_as(
+ type_=DeleteResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def list_projects(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[ProjectSortField] = None,
+ filter: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ProjectsPage]:
+ """
+ List all projects in a workspace with pagination.
+
+ Query Parameters:
+ - page, page_size: Pagination
+ - sort: Sort field
+ - filter: Advanced filters
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces/default/projects?sort=-created_at&page=1&page_size=10
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number
+
+ page_size : typing.Optional[int]
+ Items per page
+
+ sort : typing.Optional[ProjectSortField]
+ Sort field
+
+ filter : typing.Optional[str]
+ Query filter expression. Supports text and JSON syntaxes:
+ - Text: name:"value" AND status>500 with operators : ~ > >= < <= IN NOT IN AND OR and negation prefix -
+ - Object (JSON): {"name":{"$like":"value"}} with operators $eq, $like, $lt, $lte, $gt, $gte, $in, $nin, $and, $or, $not
+ - Bracket notation: ?filter[name][$like]=value
+ - Relationship traversal: ?filter[relationship][$exists]=true or ?filter[relationship][field]=value
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ProjectsPage]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/projects",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": filter,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ProjectsPage,
+ parse_obj_as(
+ type_=ProjectsPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def create_project(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[Project]:
+ """
+ Create a new project in the given workspace.
+
+ Example:
+ ```
+ POST /apis/entities/v2/workspaces/default/projects
+ {
+ "name": "ml-project",
+ "description": "Machine Learning project"
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Project name (unique within workspace). Name must start with a lowercase letter, be 2-63 characters, and contain only lowercase letters, digits, and hyphens (no consecutive hyphens, cannot end with a hyphen).
+
+ description : typing.Optional[str]
+ Optional description of the project
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[Project]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/projects",
+ method="POST",
+ json={
+ "name": name,
+ "description": description,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Project,
+ parse_obj_as(
+ type_=Project, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_project(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[Project]:
+ """
+ Get a specific project by its workspace and name.
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces/default/projects/ml-project
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[Project]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/projects/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Project,
+ parse_obj_as(
+ type_=Project, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def update_project(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[Project]:
+ """
+ Update a project's description.
+
+ Example:
+ ```
+ PUT /apis/entities/v2/workspaces/default/projects/ml-project
+ {
+ "description": "Updated description for ML project"
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ description : typing.Optional[str]
+ Updated description
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[Project]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/projects/{encode_path_param(name)}",
+ method="PUT",
+ json={
+ "description": description,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Project,
+ parse_obj_as(
+ type_=Project, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def delete_project(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[DeleteResponse]:
+ """
+ Delete a project.
+
+ Example:
+ ```
+ DELETE /apis/entities/v2/workspaces/default/projects/ml-project
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[DeleteResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/projects/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ DeleteResponse,
+ parse_obj_as(
+ type_=DeleteResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+
+class AsyncRawEntityStoreClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ async def get_entity_by_id(
+ self, id: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[Entity]:
+ """
+ Get a specific entity by its unique identifier.
+ This endpoint is primarily for debugging and internal use.
+
+ Example:
+ ```
+ GET /apis/entities/v2/entities/customization-config-5Q2LoF8z8M9JZxZsHwJKNn
+ ```
+
+ Parameters
+ ----------
+ id : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[Entity]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/entities/{encode_path_param(id)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Entity,
+ parse_obj_as(
+ type_=Entity, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def list_workspaces(
+ self,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[GenericSortField] = None,
+ filter: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[WorkspacesPage]:
+ """
+ List all workspaces with pagination.
+
+ When authentication is enabled, only workspaces the principal has access to
+ are returned. Service principals and platform admins have access to all workspaces.
+
+ Query Parameters:
+ - page, page_size: Pagination
+ - sort: Sort field
+ - filter: Advanced filters (JSON, text, or bracket notation)
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces?sort=-created_at&page=1&page_size=10
+ ```
+
+ Parameters
+ ----------
+ page : typing.Optional[int]
+ Page number
+
+ page_size : typing.Optional[int]
+ Items per page
+
+ sort : typing.Optional[GenericSortField]
+ Sort field
+
+ filter : typing.Optional[str]
+ Query filter expression. Supports text and JSON syntaxes:
+ - Text: name:"value" AND status>500 with operators : ~ > >= < <= IN NOT IN AND OR and negation prefix -
+ - Object (JSON): {"name":{"$like":"value"}} with operators $eq, $like, $lt, $lte, $gt, $gte, $in, $nin, $and, $or, $not
+ - Bracket notation: ?filter[name][$like]=value
+ - Relationship traversal: ?filter[relationship][$exists]=true or ?filter[relationship][field]=value
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[WorkspacesPage]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ "apis/entities/v2/workspaces",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": filter,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ WorkspacesPage,
+ parse_obj_as(
+ type_=WorkspacesPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def create_workspace(
+ self,
+ *,
+ name: str,
+ wait_role_propagation: typing.Optional[bool] = None,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[Workspace]:
+ """
+ Create a new workspace.
+
+ The creator is automatically granted Admin role on the workspace.
+ By default, this endpoint waits for the Admin role to propagate before returning.
+ Use `wait_role_propagation=false` to skip waiting (useful for bulk operations).
+
+ Example:
+ ```
+ POST /apis/entities/v2/workspaces
+ {
+ "name": "ml-team",
+ "description": "Machine Learning Team workspace"
+ }
+ ```
+
+ Parameters
+ ----------
+ name : str
+ Workspace name (unique identifier). Name must start with a lowercase letter, be 2-63 characters, and contain only lowercase letters, digits, and hyphens (no consecutive hyphens, cannot end with a hyphen).
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for Admin role to propagate before returning (default: true). Set to false for bulk operations.
+
+ description : typing.Optional[str]
+ Optional description of the workspace
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[Workspace]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ "apis/entities/v2/workspaces",
+ method="POST",
+ params={
+ "wait_role_propagation": wait_role_propagation,
+ },
+ json={
+ "name": name,
+ "description": description,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Workspace,
+ parse_obj_as(
+ type_=Workspace, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_workspace(
+ self, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[Workspace]:
+ """
+ Get a specific workspace by ID.
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces/ml-team
+ ```
+
+ Parameters
+ ----------
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[Workspace]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Workspace,
+ parse_obj_as(
+ type_=Workspace, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def update_workspace(
+ self,
+ name: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[Workspace]:
+ """
+ Update a workspace's description.
+
+ Example:
+ ```
+ PUT /apis/entities/v2/workspaces/ml-team
+ {
+ "description": "Updated description for ML Team"
+ }
+ ```
+
+ Parameters
+ ----------
+ name : str
+
+ description : typing.Optional[str]
+ Updated description
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[Workspace]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(name)}",
+ method="PUT",
+ json={
+ "description": description,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Workspace,
+ parse_obj_as(
+ type_=Workspace, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def delete_workspace(
+ self, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[DeleteResponse]:
+ """
+ Delete a workspace.
+
+ This marks the workspace for deletion and returns immediately. The workspace
+ will no longer be accessible via the API. An asynchronous cleanup controller
+ will handle deletion of all entities and external resources.
+
+ Role bindings are immediately deleted to revoke access.
+
+ Example:
+ ```
+ DELETE /apis/entities/v2/workspaces/ml-team
+ ```
+
+ Parameters
+ ----------
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[DeleteResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ DeleteResponse,
+ parse_obj_as(
+ type_=DeleteResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def list_entities(
+ self,
+ workspace: str,
+ entity_type: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[str] = None,
+ filter: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[EntitiesPage]:
+ """
+ List all entities of a specific type in the given workspace.
+
+ Use workspace="-" to list entities across all workspaces the principal has
+ access to.
+
+ Query Parameters:
+ - sort: Sort field
+ - page, page_size: Pagination
+ - filter: Advanced filters (JSON, text, or bracket notation)
+
+ Examples:
+ ```
+ GET /apis/entities/v2/workspaces/default/entities/customization_config?sort=-created_at
+ GET /apis/entities/v2/workspaces/-/entities/customization_config # Cross-workspace query
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ entity_type : str
+
+ page : typing.Optional[int]
+ Page number
+
+ page_size : typing.Optional[int]
+ Items per page
+
+ sort : typing.Optional[str]
+ Sort field
+
+ filter : typing.Optional[str]
+ Query filter expression. Supports text and JSON syntaxes:
+ - Text: name:"value" AND status>500 with operators : ~ > >= < <= IN NOT IN AND OR and negation prefix -
+ - Object (JSON): {"name":{"$like":"value"}} with operators $eq, $like, $lt, $lte, $gt, $gte, $in, $nin, $and, $or, $not
+ - Bracket notation: ?filter[name][$like]=value
+ - Relationship traversal: ?filter[relationship][$exists]=true or ?filter[relationship][field]=value
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[EntitiesPage]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/entities/{encode_path_param(entity_type)}",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": filter,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ EntitiesPage,
+ parse_obj_as(
+ type_=EntitiesPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def create_entity(
+ self,
+ workspace: str,
+ entity_type: str,
+ *,
+ data: typing.Dict[str, typing.Any],
+ name: typing.Optional[str] = OMIT,
+ parent: typing.Optional[str] = OMIT,
+ project: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[Entity]:
+ """
+ Create a new entity of the specified type in the given workspace.
+
+ If name is not provided, it will be auto-generated based on the entity type.
+
+ Example:
+ ```
+ POST /apis/entities/v2/workspaces/default/entities/customization_config
+ {
+ "name": "my-config",
+ "data": {
+ "target_id": "llama-2-7b",
+ "training_options": {"learning_rate": 0.01}
+ }
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ entity_type : str
+
+ data : typing.Dict[str, typing.Any]
+ Entity-specific data (schema is opaque to entity store, validated by client SDK)
+
+ name : typing.Optional[str]
+ Entity name (optional - auto-generated if not provided). Name must start with a lowercase letter, be 2-63 characters, and contain only lowercase letters, digits, and hyphens (no consecutive hyphens, cannot end with a hyphen).
+
+ parent : typing.Optional[str]
+ Parent entity ID for nested entities
+
+ project : typing.Optional[str]
+ The name of the project associated with this entity
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[Entity]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/entities/{encode_path_param(entity_type)}",
+ method="POST",
+ json={
+ "name": name,
+ "parent": parent,
+ "project": project,
+ "data": data,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Entity,
+ parse_obj_as(
+ type_=Entity, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_entity_by_name(
+ self,
+ workspace: str,
+ entity_type: str,
+ name: str,
+ *,
+ parent: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[Entity]:
+ """
+ Get a specific entity by its workspace, type, and name.
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces/default/entities/customization_config/my-config
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ entity_type : str
+
+ name : str
+
+ parent : typing.Optional[str]
+ Parent entity ID for nested entities
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[Entity]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/entities/{encode_path_param(entity_type)}/{encode_path_param(name)}",
+ method="GET",
+ params={
+ "parent": parent,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Entity,
+ parse_obj_as(
+ type_=Entity, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def update_entity_by_name(
+ self,
+ workspace: str,
+ entity_type: str,
+ name: str,
+ *,
+ data: typing.Dict[str, typing.Any],
+ parent: typing.Optional[str] = None,
+ new_name: typing.Optional[str] = OMIT,
+ project: typing.Optional[str] = OMIT,
+ expected_db_version: typing.Optional[int] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[Entity]:
+ """
+ Update an entity by its name. Optionally change the entity's name.
+
+ Example:
+ ```
+ PUT /apis/entities/v2/workspaces/default/entities/customization_config/my-config
+ {
+ "data": {
+ "target_id": "llama-2-7b",
+ "training_options": {"learning_rate": 0.02}
+ }
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ entity_type : str
+
+ name : str
+
+ data : typing.Dict[str, typing.Any]
+ Updated entity-specific data
+
+ parent : typing.Optional[str]
+ Parent entity ID for nested entities
+
+ new_name : typing.Optional[str]
+ Updated entity name (optional). Name must start with a lowercase letter, be 2-63 characters, and contain only lowercase letters, digits, and hyphens (no consecutive hyphens, cannot end with a hyphen).
+
+ project : typing.Optional[str]
+ The name of the project associated with this entity
+
+ expected_db_version : typing.Optional[int]
+ Optional database version for optimistic locking. Update only succeeds if current version matches.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[Entity]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/entities/{encode_path_param(entity_type)}/{encode_path_param(name)}",
+ method="PUT",
+ params={
+ "parent": parent,
+ },
+ json={
+ "new_name": new_name,
+ "project": project,
+ "data": data,
+ "expected_db_version": expected_db_version,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Entity,
+ parse_obj_as(
+ type_=Entity, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def delete_entity_by_name(
+ self,
+ workspace: str,
+ entity_type: str,
+ name: str,
+ *,
+ parent: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[DeleteResponse]:
+ """
+ Delete an entity by its name.
+
+ Example:
+ ```
+ DELETE /apis/entities/v2/workspaces/default/entities/customization_config/my-config
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ entity_type : str
+
+ name : str
+
+ parent : typing.Optional[str]
+ Parent entity ID for nested entities
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[DeleteResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/entities/{encode_path_param(entity_type)}/{encode_path_param(name)}",
+ method="DELETE",
+ params={
+ "parent": parent,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ DeleteResponse,
+ parse_obj_as(
+ type_=DeleteResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def list_workspace_members(
+ self, workspace: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[WorkspaceMemberListResponse]:
+ """
+ List all members of a workspace with their roles.
+
+ Returns a list of all principals with active role bindings in the workspace.
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces/ml-team/members
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[WorkspaceMemberListResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/members",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ WorkspaceMemberListResponse,
+ parse_obj_as(
+ type_=WorkspaceMemberListResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def add_workspace_member(
+ self,
+ workspace: str,
+ *,
+ principal: str,
+ wait_role_propagation: typing.Optional[bool] = None,
+ roles: typing.Optional[typing.Sequence[str]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[WorkspaceMember]:
+ """
+ Add a new member to the workspace with specified roles.
+
+ This creates role bindings for the specified principal with the given roles.
+ By default, this endpoint waits for the roles to propagate before returning.
+ Use `wait_role_propagation=false` to skip waiting (useful for bulk operations).
+
+ Example:
+ ```
+ POST /apis/entities/v2/workspaces/ml-team/members
+ {
+ "principal": "user@example.com",
+ "roles": ["Editor"]
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ principal : str
+ The principal identifier (email, user ID, or group ID)
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for roles to propagate before returning (default: true). Set to false for bulk operations.
+
+ roles : typing.Optional[typing.Sequence[str]]
+ List of roles to grant to the principal
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[WorkspaceMember]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/members",
+ method="POST",
+ params={
+ "wait_role_propagation": wait_role_propagation,
+ },
+ json={
+ "principal": principal,
+ "roles": roles,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ WorkspaceMember,
+ parse_obj_as(
+ type_=WorkspaceMember, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def update_workspace_member(
+ self,
+ workspace: str,
+ principal_id: str,
+ *,
+ roles: typing.Sequence[str],
+ wait_role_propagation: typing.Optional[bool] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[WorkspaceMember]:
+ """
+ Update the roles for a workspace member.
+
+ This will revoke existing roles not in the new list and add new roles.
+ By default, this endpoint waits for the roles to propagate before returning.
+ Use `wait_role_propagation=false` to skip waiting (useful for bulk operations).
+
+ Example:
+ ```
+ PUT /apis/entities/v2/workspaces/ml-team/members/user@example.com
+ {
+ "roles": ["Viewer", "Editor"]
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ principal_id : str
+
+ roles : typing.Sequence[str]
+ Updated list of roles for the principal
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for roles to propagate before returning (default: true). Set to false for bulk operations.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[WorkspaceMember]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/members/{encode_path_param(principal_id)}",
+ method="PUT",
+ params={
+ "wait_role_propagation": wait_role_propagation,
+ },
+ json={
+ "roles": roles,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ WorkspaceMember,
+ parse_obj_as(
+ type_=WorkspaceMember, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def remove_workspace_member(
+ self,
+ workspace: str,
+ principal_id: str,
+ *,
+ wait_role_propagation: typing.Optional[bool] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[DeleteResponse]:
+ """
+ Remove a member from the workspace by revoking all their roles.
+
+ This revokes all active role bindings for the principal in the workspace.
+ By default, this endpoint waits for all roles to be revoked before returning.
+ Use `wait_role_propagation=false` to skip waiting (useful for bulk operations).
+
+ Example:
+ ```
+ DELETE /apis/entities/v2/workspaces/ml-team/members/user@example.com
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ principal_id : str
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for roles to propagate before returning (default: true). Set to false for bulk operations.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[DeleteResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/members/{encode_path_param(principal_id)}",
+ method="DELETE",
+ params={
+ "wait_role_propagation": wait_role_propagation,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ DeleteResponse,
+ parse_obj_as(
+ type_=DeleteResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def list_projects(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[ProjectSortField] = None,
+ filter: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ProjectsPage]:
+ """
+ List all projects in a workspace with pagination.
+
+ Query Parameters:
+ - page, page_size: Pagination
+ - sort: Sort field
+ - filter: Advanced filters
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces/default/projects?sort=-created_at&page=1&page_size=10
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number
+
+ page_size : typing.Optional[int]
+ Items per page
+
+ sort : typing.Optional[ProjectSortField]
+ Sort field
+
+ filter : typing.Optional[str]
+ Query filter expression. Supports text and JSON syntaxes:
+ - Text: name:"value" AND status>500 with operators : ~ > >= < <= IN NOT IN AND OR and negation prefix -
+ - Object (JSON): {"name":{"$like":"value"}} with operators $eq, $like, $lt, $lte, $gt, $gte, $in, $nin, $and, $or, $not
+ - Bracket notation: ?filter[name][$like]=value
+ - Relationship traversal: ?filter[relationship][$exists]=true or ?filter[relationship][field]=value
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ProjectsPage]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/projects",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": filter,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ProjectsPage,
+ parse_obj_as(
+ type_=ProjectsPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def create_project(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[Project]:
+ """
+ Create a new project in the given workspace.
+
+ Example:
+ ```
+ POST /apis/entities/v2/workspaces/default/projects
+ {
+ "name": "ml-project",
+ "description": "Machine Learning project"
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Project name (unique within workspace). Name must start with a lowercase letter, be 2-63 characters, and contain only lowercase letters, digits, and hyphens (no consecutive hyphens, cannot end with a hyphen).
+
+ description : typing.Optional[str]
+ Optional description of the project
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[Project]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/projects",
+ method="POST",
+ json={
+ "name": name,
+ "description": description,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Project,
+ parse_obj_as(
+ type_=Project, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_project(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[Project]:
+ """
+ Get a specific project by its workspace and name.
+
+ Example:
+ ```
+ GET /apis/entities/v2/workspaces/default/projects/ml-project
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[Project]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/projects/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Project,
+ parse_obj_as(
+ type_=Project, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def update_project(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[Project]:
+ """
+ Update a project's description.
+
+ Example:
+ ```
+ PUT /apis/entities/v2/workspaces/default/projects/ml-project
+ {
+ "description": "Updated description for ML project"
+ }
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ description : typing.Optional[str]
+ Updated description
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[Project]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/projects/{encode_path_param(name)}",
+ method="PUT",
+ json={
+ "description": description,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Project,
+ parse_obj_as(
+ type_=Project, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def delete_project(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[DeleteResponse]:
+ """
+ Delete a project.
+
+ Example:
+ ```
+ DELETE /apis/entities/v2/workspaces/default/projects/ml-project
+ ```
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[DeleteResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/entities/v2/workspaces/{encode_path_param(workspace)}/projects/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ DeleteResponse,
+ parse_obj_as(
+ type_=DeleteResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
diff --git a/sdks/python/errors/__init__.py b/sdks/python/errors/__init__.py
new file mode 100644
index 0000000000..6003ceb8b8
--- /dev/null
+++ b/sdks/python/errors/__init__.py
@@ -0,0 +1,53 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
+import typing
+from importlib import import_module
+
+if typing.TYPE_CHECKING:
+ from .bad_request_error import BadRequestError
+ from .conflict_error import ConflictError
+ from .internal_server_error import InternalServerError
+ from .not_found_error import NotFoundError
+ from .service_unavailable_error import ServiceUnavailableError
+ from .unprocessable_entity_error import UnprocessableEntityError
+_dynamic_imports: typing.Dict[str, str] = {
+ "BadRequestError": ".bad_request_error",
+ "ConflictError": ".conflict_error",
+ "InternalServerError": ".internal_server_error",
+ "NotFoundError": ".not_found_error",
+ "ServiceUnavailableError": ".service_unavailable_error",
+ "UnprocessableEntityError": ".unprocessable_entity_error",
+}
+
+
+def __getattr__(attr_name: str) -> typing.Any:
+ module_name = _dynamic_imports.get(attr_name)
+ if module_name is None:
+ raise AttributeError(f"No {attr_name} found in _dynamic_imports for module name -> {__name__}")
+ try:
+ module = import_module(module_name, __package__)
+ if module_name == f".{attr_name}":
+ return module
+ else:
+ return getattr(module, attr_name)
+ except ImportError as e:
+ raise ImportError(f"Failed to import {attr_name} from {module_name}: {e}") from e
+ except AttributeError as e:
+ raise AttributeError(f"Failed to get {attr_name} from {module_name}: {e}") from e
+
+
+def __dir__():
+ lazy_attrs = list(_dynamic_imports.keys())
+ return sorted(lazy_attrs)
+
+
+__all__ = [
+ "BadRequestError",
+ "ConflictError",
+ "InternalServerError",
+ "NotFoundError",
+ "ServiceUnavailableError",
+ "UnprocessableEntityError",
+]
diff --git a/sdks/python/errors/bad_request_error.py b/sdks/python/errors/bad_request_error.py
new file mode 100644
index 0000000000..ec78e2697a
--- /dev/null
+++ b/sdks/python/errors/bad_request_error.py
@@ -0,0 +1,10 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.api_error import ApiError
+
+
+class BadRequestError(ApiError):
+ def __init__(self, body: typing.Any, headers: typing.Optional[typing.Dict[str, str]] = None):
+ super().__init__(status_code=400, headers=headers, body=body)
diff --git a/sdks/python/errors/conflict_error.py b/sdks/python/errors/conflict_error.py
new file mode 100644
index 0000000000..be04e01aee
--- /dev/null
+++ b/sdks/python/errors/conflict_error.py
@@ -0,0 +1,10 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.api_error import ApiError
+
+
+class ConflictError(ApiError):
+ def __init__(self, body: typing.Any, headers: typing.Optional[typing.Dict[str, str]] = None):
+ super().__init__(status_code=409, headers=headers, body=body)
diff --git a/sdks/python/errors/internal_server_error.py b/sdks/python/errors/internal_server_error.py
new file mode 100644
index 0000000000..fabcc45cf7
--- /dev/null
+++ b/sdks/python/errors/internal_server_error.py
@@ -0,0 +1,10 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.api_error import ApiError
+
+
+class InternalServerError(ApiError):
+ def __init__(self, body: typing.Any, headers: typing.Optional[typing.Dict[str, str]] = None):
+ super().__init__(status_code=500, headers=headers, body=body)
diff --git a/sdks/python/errors/not_found_error.py b/sdks/python/errors/not_found_error.py
new file mode 100644
index 0000000000..75f557df64
--- /dev/null
+++ b/sdks/python/errors/not_found_error.py
@@ -0,0 +1,10 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.api_error import ApiError
+
+
+class NotFoundError(ApiError):
+ def __init__(self, body: typing.Any, headers: typing.Optional[typing.Dict[str, str]] = None):
+ super().__init__(status_code=404, headers=headers, body=body)
diff --git a/sdks/python/errors/service_unavailable_error.py b/sdks/python/errors/service_unavailable_error.py
new file mode 100644
index 0000000000..950c474028
--- /dev/null
+++ b/sdks/python/errors/service_unavailable_error.py
@@ -0,0 +1,10 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.api_error import ApiError
+
+
+class ServiceUnavailableError(ApiError):
+ def __init__(self, body: typing.Any, headers: typing.Optional[typing.Dict[str, str]] = None):
+ super().__init__(status_code=503, headers=headers, body=body)
diff --git a/sdks/python/errors/unprocessable_entity_error.py b/sdks/python/errors/unprocessable_entity_error.py
new file mode 100644
index 0000000000..1c801a4bc0
--- /dev/null
+++ b/sdks/python/errors/unprocessable_entity_error.py
@@ -0,0 +1,10 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.api_error import ApiError
+
+
+class UnprocessableEntityError(ApiError):
+ def __init__(self, body: typing.Any, headers: typing.Optional[typing.Dict[str, str]] = None):
+ super().__init__(status_code=422, headers=headers, body=body)
diff --git a/sdks/python/evaluator_results/__init__.py b/sdks/python/evaluator_results/__init__.py
new file mode 100644
index 0000000000..5cde0202dc
--- /dev/null
+++ b/sdks/python/evaluator_results/__init__.py
@@ -0,0 +1,4 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
diff --git a/sdks/python/evaluator_results/client.py b/sdks/python/evaluator_results/client.py
new file mode 100644
index 0000000000..8fb0814dec
--- /dev/null
+++ b/sdks/python/evaluator_results/client.py
@@ -0,0 +1,469 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.request_options import RequestOptions
+from ..types.evaluator_result import EvaluatorResult
+from ..types.evaluator_result_data_type import EvaluatorResultDataType
+from ..types.evaluator_result_filter import EvaluatorResultFilter
+from ..types.evaluator_result_sort_field import EvaluatorResultSortField
+from ..types.evaluator_results_page import EvaluatorResultsPage
+from .raw_client import AsyncRawEvaluatorResultsClient, RawEvaluatorResultsClient
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class EvaluatorResultsClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._raw_client = RawEvaluatorResultsClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> RawEvaluatorResultsClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ RawEvaluatorResultsClient
+ """
+ return self._raw_client
+
+ def list_evaluator_results(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[EvaluatorResultSortField] = None,
+ filter: typing.Optional[EvaluatorResultFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> EvaluatorResultsPage:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[EvaluatorResultSortField]
+
+ filter : typing.Optional[EvaluatorResultFilter]
+ Filter evaluator results by span_id, session_id, name, data_type, created_by, value range, and created_at range.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ EvaluatorResultsPage
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.evaluator_results.list_evaluator_results(
+ workspace="workspace",
+ )
+ """
+ _response = self._raw_client.list_evaluator_results(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ def create_evaluator_result(
+ self,
+ workspace: str,
+ *,
+ span_id: str,
+ session_id: str,
+ name: str,
+ data_type: EvaluatorResultDataType,
+ value: typing.Optional[float] = OMIT,
+ string_value: typing.Optional[str] = OMIT,
+ comment: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> EvaluatorResult:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ span_id : str
+ Target span id. Not validated against existing spans (loose target policy).
+
+ session_id : str
+ Session id the target span belongs to. Denormalized so session-scoped reads stay fast.
+
+ name : str
+ Evaluator / metric identity (e.g. 'faithfulness/v1').
+
+ data_type : EvaluatorResultDataType
+ Discriminator for which of value / string_value carries the payload.
+
+ value : typing.Optional[float]
+ Numeric value. Required when data_type is NUMERIC or BOOLEAN (0|1).
+
+ string_value : typing.Optional[str]
+ String value. Required when data_type is CATEGORICAL or TEXT.
+
+ comment : typing.Optional[str]
+ Free-text rationale or explanation.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ EvaluatorResult
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.evaluator_results.create_evaluator_result(
+ workspace="workspace",
+ span_id="span_id",
+ session_id="session_id",
+ name="name",
+ data_type="NUMERIC",
+ )
+ """
+ _response = self._raw_client.create_evaluator_result(
+ workspace,
+ span_id=span_id,
+ session_id=session_id,
+ name=name,
+ data_type=data_type,
+ value=value,
+ string_value=string_value,
+ comment=comment,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def get_evaluator_result(
+ self, workspace: str, evaluator_result_id: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> EvaluatorResult:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ evaluator_result_id : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ EvaluatorResult
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.evaluator_results.get_evaluator_result(
+ workspace="workspace",
+ evaluator_result_id="evaluator_result_id",
+ )
+ """
+ _response = self._raw_client.get_evaluator_result(
+ workspace, evaluator_result_id, request_options=request_options
+ )
+ return _response.data
+
+ def list_evaluator_results_for_span(
+ self, workspace: str, span_id: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.List[EvaluatorResult]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ span_id : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.List[EvaluatorResult]
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.evaluator_results.list_evaluator_results_for_span(
+ workspace="workspace",
+ span_id="span_id",
+ )
+ """
+ _response = self._raw_client.list_evaluator_results_for_span(
+ workspace, span_id, request_options=request_options
+ )
+ return _response.data
+
+
+class AsyncEvaluatorResultsClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._raw_client = AsyncRawEvaluatorResultsClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> AsyncRawEvaluatorResultsClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ AsyncRawEvaluatorResultsClient
+ """
+ return self._raw_client
+
+ async def list_evaluator_results(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[EvaluatorResultSortField] = None,
+ filter: typing.Optional[EvaluatorResultFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> EvaluatorResultsPage:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[EvaluatorResultSortField]
+
+ filter : typing.Optional[EvaluatorResultFilter]
+ Filter evaluator results by span_id, session_id, name, data_type, created_by, value range, and created_at range.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ EvaluatorResultsPage
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.evaluator_results.list_evaluator_results(
+ workspace="workspace",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_evaluator_results(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ async def create_evaluator_result(
+ self,
+ workspace: str,
+ *,
+ span_id: str,
+ session_id: str,
+ name: str,
+ data_type: EvaluatorResultDataType,
+ value: typing.Optional[float] = OMIT,
+ string_value: typing.Optional[str] = OMIT,
+ comment: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> EvaluatorResult:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ span_id : str
+ Target span id. Not validated against existing spans (loose target policy).
+
+ session_id : str
+ Session id the target span belongs to. Denormalized so session-scoped reads stay fast.
+
+ name : str
+ Evaluator / metric identity (e.g. 'faithfulness/v1').
+
+ data_type : EvaluatorResultDataType
+ Discriminator for which of value / string_value carries the payload.
+
+ value : typing.Optional[float]
+ Numeric value. Required when data_type is NUMERIC or BOOLEAN (0|1).
+
+ string_value : typing.Optional[str]
+ String value. Required when data_type is CATEGORICAL or TEXT.
+
+ comment : typing.Optional[str]
+ Free-text rationale or explanation.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ EvaluatorResult
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.evaluator_results.create_evaluator_result(
+ workspace="workspace",
+ span_id="span_id",
+ session_id="session_id",
+ name="name",
+ data_type="NUMERIC",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.create_evaluator_result(
+ workspace,
+ span_id=span_id,
+ session_id=session_id,
+ name=name,
+ data_type=data_type,
+ value=value,
+ string_value=string_value,
+ comment=comment,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def get_evaluator_result(
+ self, workspace: str, evaluator_result_id: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> EvaluatorResult:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ evaluator_result_id : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ EvaluatorResult
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.evaluator_results.get_evaluator_result(
+ workspace="workspace",
+ evaluator_result_id="evaluator_result_id",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_evaluator_result(
+ workspace, evaluator_result_id, request_options=request_options
+ )
+ return _response.data
+
+ async def list_evaluator_results_for_span(
+ self, workspace: str, span_id: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.List[EvaluatorResult]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ span_id : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.List[EvaluatorResult]
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.evaluator_results.list_evaluator_results_for_span(
+ workspace="workspace",
+ span_id="span_id",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_evaluator_results_for_span(
+ workspace, span_id, request_options=request_options
+ )
+ return _response.data
diff --git a/sdks/python/evaluator_results/raw_client.py b/sdks/python/evaluator_results/raw_client.py
new file mode 100644
index 0000000000..c4a64f6bb3
--- /dev/null
+++ b/sdks/python/evaluator_results/raw_client.py
@@ -0,0 +1,589 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+from json.decoder import JSONDecodeError
+
+from ..core.api_error import ApiError
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.http_response import AsyncHttpResponse, HttpResponse
+from ..core.jsonable_encoder import encode_path_param
+from ..core.parse_error import ParsingError
+from ..core.pydantic_utilities import parse_obj_as
+from ..core.request_options import RequestOptions
+from ..core.serialization import convert_and_respect_annotation_metadata
+from ..errors.unprocessable_entity_error import UnprocessableEntityError
+from ..types.evaluator_result import EvaluatorResult
+from ..types.evaluator_result_data_type import EvaluatorResultDataType
+from ..types.evaluator_result_filter import EvaluatorResultFilter
+from ..types.evaluator_result_sort_field import EvaluatorResultSortField
+from ..types.evaluator_results_page import EvaluatorResultsPage
+from pydantic import ValidationError
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class RawEvaluatorResultsClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ def list_evaluator_results(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[EvaluatorResultSortField] = None,
+ filter: typing.Optional[EvaluatorResultFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[EvaluatorResultsPage]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[EvaluatorResultSortField]
+
+ filter : typing.Optional[EvaluatorResultFilter]
+ Filter evaluator results by span_id, session_id, name, data_type, created_by, value range, and created_at range.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[EvaluatorResultsPage]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/evaluator-results",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=EvaluatorResultFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ EvaluatorResultsPage,
+ parse_obj_as(
+ type_=EvaluatorResultsPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def create_evaluator_result(
+ self,
+ workspace: str,
+ *,
+ span_id: str,
+ session_id: str,
+ name: str,
+ data_type: EvaluatorResultDataType,
+ value: typing.Optional[float] = OMIT,
+ string_value: typing.Optional[str] = OMIT,
+ comment: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[EvaluatorResult]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ span_id : str
+ Target span id. Not validated against existing spans (loose target policy).
+
+ session_id : str
+ Session id the target span belongs to. Denormalized so session-scoped reads stay fast.
+
+ name : str
+ Evaluator / metric identity (e.g. 'faithfulness/v1').
+
+ data_type : EvaluatorResultDataType
+ Discriminator for which of value / string_value carries the payload.
+
+ value : typing.Optional[float]
+ Numeric value. Required when data_type is NUMERIC or BOOLEAN (0|1).
+
+ string_value : typing.Optional[str]
+ String value. Required when data_type is CATEGORICAL or TEXT.
+
+ comment : typing.Optional[str]
+ Free-text rationale or explanation.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[EvaluatorResult]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/evaluator-results",
+ method="POST",
+ json={
+ "span_id": span_id,
+ "session_id": session_id,
+ "name": name,
+ "value": value,
+ "string_value": string_value,
+ "data_type": data_type,
+ "comment": comment,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ EvaluatorResult,
+ parse_obj_as(
+ type_=EvaluatorResult, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_evaluator_result(
+ self, workspace: str, evaluator_result_id: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[EvaluatorResult]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ evaluator_result_id : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[EvaluatorResult]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/evaluator-results/{encode_path_param(evaluator_result_id)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ EvaluatorResult,
+ parse_obj_as(
+ type_=EvaluatorResult, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def list_evaluator_results_for_span(
+ self, workspace: str, span_id: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[typing.List[EvaluatorResult]]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ span_id : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.List[EvaluatorResult]]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/spans/{encode_path_param(span_id)}/evaluator-results",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.List[EvaluatorResult],
+ parse_obj_as(
+ type_=typing.List[EvaluatorResult], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+
+class AsyncRawEvaluatorResultsClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ async def list_evaluator_results(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[EvaluatorResultSortField] = None,
+ filter: typing.Optional[EvaluatorResultFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[EvaluatorResultsPage]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[EvaluatorResultSortField]
+
+ filter : typing.Optional[EvaluatorResultFilter]
+ Filter evaluator results by span_id, session_id, name, data_type, created_by, value range, and created_at range.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[EvaluatorResultsPage]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/evaluator-results",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=EvaluatorResultFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ EvaluatorResultsPage,
+ parse_obj_as(
+ type_=EvaluatorResultsPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def create_evaluator_result(
+ self,
+ workspace: str,
+ *,
+ span_id: str,
+ session_id: str,
+ name: str,
+ data_type: EvaluatorResultDataType,
+ value: typing.Optional[float] = OMIT,
+ string_value: typing.Optional[str] = OMIT,
+ comment: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[EvaluatorResult]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ span_id : str
+ Target span id. Not validated against existing spans (loose target policy).
+
+ session_id : str
+ Session id the target span belongs to. Denormalized so session-scoped reads stay fast.
+
+ name : str
+ Evaluator / metric identity (e.g. 'faithfulness/v1').
+
+ data_type : EvaluatorResultDataType
+ Discriminator for which of value / string_value carries the payload.
+
+ value : typing.Optional[float]
+ Numeric value. Required when data_type is NUMERIC or BOOLEAN (0|1).
+
+ string_value : typing.Optional[str]
+ String value. Required when data_type is CATEGORICAL or TEXT.
+
+ comment : typing.Optional[str]
+ Free-text rationale or explanation.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[EvaluatorResult]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/evaluator-results",
+ method="POST",
+ json={
+ "span_id": span_id,
+ "session_id": session_id,
+ "name": name,
+ "value": value,
+ "string_value": string_value,
+ "data_type": data_type,
+ "comment": comment,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ EvaluatorResult,
+ parse_obj_as(
+ type_=EvaluatorResult, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_evaluator_result(
+ self, workspace: str, evaluator_result_id: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[EvaluatorResult]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ evaluator_result_id : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[EvaluatorResult]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/evaluator-results/{encode_path_param(evaluator_result_id)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ EvaluatorResult,
+ parse_obj_as(
+ type_=EvaluatorResult, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def list_evaluator_results_for_span(
+ self, workspace: str, span_id: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[typing.List[EvaluatorResult]]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ span_id : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.List[EvaluatorResult]]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/spans/{encode_path_param(span_id)}/evaluator-results",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.List[EvaluatorResult],
+ parse_obj_as(
+ type_=typing.List[EvaluatorResult], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
diff --git a/sdks/python/experiment_groups/__init__.py b/sdks/python/experiment_groups/__init__.py
new file mode 100644
index 0000000000..2a2adb83a1
--- /dev/null
+++ b/sdks/python/experiment_groups/__init__.py
@@ -0,0 +1,36 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
+import typing
+from importlib import import_module
+
+if typing.TYPE_CHECKING:
+ from .types import ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort
+_dynamic_imports: typing.Dict[str, str] = {
+ "ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort": ".types"
+}
+
+
+def __getattr__(attr_name: str) -> typing.Any:
+ module_name = _dynamic_imports.get(attr_name)
+ if module_name is None:
+ raise AttributeError(f"No {attr_name} found in _dynamic_imports for module name -> {__name__}")
+ try:
+ module = import_module(module_name, __package__)
+ if module_name == f".{attr_name}":
+ return module
+ else:
+ return getattr(module, attr_name)
+ except ImportError as e:
+ raise ImportError(f"Failed to import {attr_name} from {module_name}: {e}") from e
+ except AttributeError as e:
+ raise AttributeError(f"Failed to get {attr_name} from {module_name}: {e}") from e
+
+
+def __dir__():
+ lazy_attrs = list(_dynamic_imports.keys())
+ return sorted(lazy_attrs)
+
+
+__all__ = ["ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort"]
diff --git a/sdks/python/experiment_groups/client.py b/sdks/python/experiment_groups/client.py
new file mode 100644
index 0000000000..7c0e8dbf8f
--- /dev/null
+++ b/sdks/python/experiment_groups/client.py
@@ -0,0 +1,504 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.request_options import RequestOptions
+from ..types.experiment_group_filter import ExperimentGroupFilter
+from ..types.experiment_group_response import ExperimentGroupResponse
+from ..types.experiment_group_responses_page import ExperimentGroupResponsesPage
+from .raw_client import AsyncRawExperimentGroupsClient, RawExperimentGroupsClient
+from .types.list_experiment_groups_apis_intake_v2workspaces_workspace_experiment_groups_get_request_sort import (
+ ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort,
+)
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class ExperimentGroupsClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._raw_client = RawExperimentGroupsClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> RawExperimentGroupsClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ RawExperimentGroupsClient
+ """
+ return self._raw_client
+
+ def list_experiment_groups(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort] = None,
+ filter: typing.Optional[ExperimentGroupFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ExperimentGroupResponsesPage:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort]
+ Sort field; prefix with '-' for descending.
+
+ filter : typing.Optional[ExperimentGroupFilter]
+ Filter experiment groups by name.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ExperimentGroupResponsesPage
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.experiment_groups.list_experiment_groups(
+ workspace="workspace",
+ )
+ """
+ _response = self._raw_client.list_experiment_groups(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ def create_experiment_group(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ExperimentGroupResponse:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Workspace-unique group name.
+
+ description : typing.Optional[str]
+ Human-readable purpose of the group.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ExperimentGroupResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.experiment_groups.create_experiment_group(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.create_experiment_group(
+ workspace, name=name, description=description, request_options=request_options
+ )
+ return _response.data
+
+ def get_experiment_group(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> ExperimentGroupResponse:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ExperimentGroupResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.experiment_groups.get_experiment_group(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.get_experiment_group(workspace, name, request_options=request_options)
+ return _response.data
+
+ def update_experiment_group(
+ self,
+ workspace: str,
+ name_: str,
+ *,
+ name: str,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ExperimentGroupResponse:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name_ : str
+
+ name : str
+ Workspace-unique group name.
+
+ description : typing.Optional[str]
+ Human-readable purpose of the group.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ExperimentGroupResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.experiment_groups.update_experiment_group(
+ workspace="workspace",
+ name_="name",
+ name="name",
+ )
+ """
+ _response = self._raw_client.update_experiment_group(
+ workspace, name_, name=name, description=description, request_options=request_options
+ )
+ return _response.data
+
+ def delete_experiment_group(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> None:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.experiment_groups.delete_experiment_group(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.delete_experiment_group(workspace, name, request_options=request_options)
+ return _response.data
+
+
+class AsyncExperimentGroupsClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._raw_client = AsyncRawExperimentGroupsClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> AsyncRawExperimentGroupsClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ AsyncRawExperimentGroupsClient
+ """
+ return self._raw_client
+
+ async def list_experiment_groups(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort] = None,
+ filter: typing.Optional[ExperimentGroupFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ExperimentGroupResponsesPage:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort]
+ Sort field; prefix with '-' for descending.
+
+ filter : typing.Optional[ExperimentGroupFilter]
+ Filter experiment groups by name.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ExperimentGroupResponsesPage
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.experiment_groups.list_experiment_groups(
+ workspace="workspace",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_experiment_groups(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ async def create_experiment_group(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ExperimentGroupResponse:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Workspace-unique group name.
+
+ description : typing.Optional[str]
+ Human-readable purpose of the group.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ExperimentGroupResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.experiment_groups.create_experiment_group(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.create_experiment_group(
+ workspace, name=name, description=description, request_options=request_options
+ )
+ return _response.data
+
+ async def get_experiment_group(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> ExperimentGroupResponse:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ExperimentGroupResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.experiment_groups.get_experiment_group(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_experiment_group(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def update_experiment_group(
+ self,
+ workspace: str,
+ name_: str,
+ *,
+ name: str,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ExperimentGroupResponse:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name_ : str
+
+ name : str
+ Workspace-unique group name.
+
+ description : typing.Optional[str]
+ Human-readable purpose of the group.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ExperimentGroupResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.experiment_groups.update_experiment_group(
+ workspace="workspace",
+ name_="name",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.update_experiment_group(
+ workspace, name_, name=name, description=description, request_options=request_options
+ )
+ return _response.data
+
+ async def delete_experiment_group(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> None:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.experiment_groups.delete_experiment_group(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.delete_experiment_group(workspace, name, request_options=request_options)
+ return _response.data
diff --git a/sdks/python/experiment_groups/raw_client.py b/sdks/python/experiment_groups/raw_client.py
new file mode 100644
index 0000000000..93f683b78a
--- /dev/null
+++ b/sdks/python/experiment_groups/raw_client.py
@@ -0,0 +1,784 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+from json.decoder import JSONDecodeError
+
+from ..core.api_error import ApiError
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.http_response import AsyncHttpResponse, HttpResponse
+from ..core.jsonable_encoder import encode_path_param
+from ..core.parse_error import ParsingError
+from ..core.pydantic_utilities import parse_obj_as
+from ..core.request_options import RequestOptions
+from ..core.serialization import convert_and_respect_annotation_metadata
+from ..errors.conflict_error import ConflictError
+from ..errors.not_found_error import NotFoundError
+from ..errors.unprocessable_entity_error import UnprocessableEntityError
+from ..types.experiment_group_filter import ExperimentGroupFilter
+from ..types.experiment_group_response import ExperimentGroupResponse
+from ..types.experiment_group_responses_page import ExperimentGroupResponsesPage
+from .types.list_experiment_groups_apis_intake_v2workspaces_workspace_experiment_groups_get_request_sort import (
+ ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort,
+)
+from pydantic import ValidationError
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class RawExperimentGroupsClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ def list_experiment_groups(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort] = None,
+ filter: typing.Optional[ExperimentGroupFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ExperimentGroupResponsesPage]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort]
+ Sort field; prefix with '-' for descending.
+
+ filter : typing.Optional[ExperimentGroupFilter]
+ Filter experiment groups by name.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ExperimentGroupResponsesPage]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiment-groups",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=ExperimentGroupFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ExperimentGroupResponsesPage,
+ parse_obj_as(
+ type_=ExperimentGroupResponsesPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def create_experiment_group(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ExperimentGroupResponse]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Workspace-unique group name.
+
+ description : typing.Optional[str]
+ Human-readable purpose of the group.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ExperimentGroupResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiment-groups",
+ method="POST",
+ json={
+ "name": name,
+ "description": description,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ExperimentGroupResponse,
+ parse_obj_as(
+ type_=ExperimentGroupResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 409:
+ raise ConflictError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_experiment_group(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[ExperimentGroupResponse]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ExperimentGroupResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiment-groups/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ExperimentGroupResponse,
+ parse_obj_as(
+ type_=ExperimentGroupResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def update_experiment_group(
+ self,
+ workspace: str,
+ name_: str,
+ *,
+ name: str,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ExperimentGroupResponse]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name_ : str
+
+ name : str
+ Workspace-unique group name.
+
+ description : typing.Optional[str]
+ Human-readable purpose of the group.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ExperimentGroupResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiment-groups/{encode_path_param(name_)}",
+ method="PUT",
+ json={
+ "name": name,
+ "description": description,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ExperimentGroupResponse,
+ parse_obj_as(
+ type_=ExperimentGroupResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 409:
+ raise ConflictError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def delete_experiment_group(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[None]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[None]
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiment-groups/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return HttpResponse(response=_response, data=None)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+
+class AsyncRawExperimentGroupsClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ async def list_experiment_groups(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort] = None,
+ filter: typing.Optional[ExperimentGroupFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ExperimentGroupResponsesPage]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort]
+ Sort field; prefix with '-' for descending.
+
+ filter : typing.Optional[ExperimentGroupFilter]
+ Filter experiment groups by name.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ExperimentGroupResponsesPage]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiment-groups",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=ExperimentGroupFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ExperimentGroupResponsesPage,
+ parse_obj_as(
+ type_=ExperimentGroupResponsesPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def create_experiment_group(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ExperimentGroupResponse]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Workspace-unique group name.
+
+ description : typing.Optional[str]
+ Human-readable purpose of the group.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ExperimentGroupResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiment-groups",
+ method="POST",
+ json={
+ "name": name,
+ "description": description,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ExperimentGroupResponse,
+ parse_obj_as(
+ type_=ExperimentGroupResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 409:
+ raise ConflictError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_experiment_group(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[ExperimentGroupResponse]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ExperimentGroupResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiment-groups/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ExperimentGroupResponse,
+ parse_obj_as(
+ type_=ExperimentGroupResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def update_experiment_group(
+ self,
+ workspace: str,
+ name_: str,
+ *,
+ name: str,
+ description: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ExperimentGroupResponse]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name_ : str
+
+ name : str
+ Workspace-unique group name.
+
+ description : typing.Optional[str]
+ Human-readable purpose of the group.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ExperimentGroupResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiment-groups/{encode_path_param(name_)}",
+ method="PUT",
+ json={
+ "name": name,
+ "description": description,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ExperimentGroupResponse,
+ parse_obj_as(
+ type_=ExperimentGroupResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 409:
+ raise ConflictError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def delete_experiment_group(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[None]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[None]
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiment-groups/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return AsyncHttpResponse(response=_response, data=None)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
diff --git a/sdks/python/experiment_groups/types/__init__.py b/sdks/python/experiment_groups/types/__init__.py
new file mode 100644
index 0000000000..b96b5f8d6e
--- /dev/null
+++ b/sdks/python/experiment_groups/types/__init__.py
@@ -0,0 +1,38 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
+import typing
+from importlib import import_module
+
+if typing.TYPE_CHECKING:
+ from .list_experiment_groups_apis_intake_v2workspaces_workspace_experiment_groups_get_request_sort import (
+ ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort,
+ )
+_dynamic_imports: typing.Dict[str, str] = {
+ "ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort": ".list_experiment_groups_apis_intake_v2workspaces_workspace_experiment_groups_get_request_sort"
+}
+
+
+def __getattr__(attr_name: str) -> typing.Any:
+ module_name = _dynamic_imports.get(attr_name)
+ if module_name is None:
+ raise AttributeError(f"No {attr_name} found in _dynamic_imports for module name -> {__name__}")
+ try:
+ module = import_module(module_name, __package__)
+ if module_name == f".{attr_name}":
+ return module
+ else:
+ return getattr(module, attr_name)
+ except ImportError as e:
+ raise ImportError(f"Failed to import {attr_name} from {module_name}: {e}") from e
+ except AttributeError as e:
+ raise AttributeError(f"Failed to get {attr_name} from {module_name}: {e}") from e
+
+
+def __dir__():
+ lazy_attrs = list(_dynamic_imports.keys())
+ return sorted(lazy_attrs)
+
+
+__all__ = ["ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort"]
diff --git a/sdks/python/experiment_groups/types/list_experiment_groups_apis_intake_v2workspaces_workspace_experiment_groups_get_request_sort.py b/sdks/python/experiment_groups/types/list_experiment_groups_apis_intake_v2workspaces_workspace_experiment_groups_get_request_sort.py
new file mode 100644
index 0000000000..b8589f6926
--- /dev/null
+++ b/sdks/python/experiment_groups/types/list_experiment_groups_apis_intake_v2workspaces_workspace_experiment_groups_get_request_sort.py
@@ -0,0 +1,7 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort = typing.Union[
+ typing.Literal["-created_at", "created_at", "-updated_at", "updated_at", "-name"], typing.Any
+]
diff --git a/sdks/python/experiments/__init__.py b/sdks/python/experiments/__init__.py
new file mode 100644
index 0000000000..4a09452243
--- /dev/null
+++ b/sdks/python/experiments/__init__.py
@@ -0,0 +1,36 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
+import typing
+from importlib import import_module
+
+if typing.TYPE_CHECKING:
+ from .types import ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort
+_dynamic_imports: typing.Dict[str, str] = {
+ "ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort": ".types"
+}
+
+
+def __getattr__(attr_name: str) -> typing.Any:
+ module_name = _dynamic_imports.get(attr_name)
+ if module_name is None:
+ raise AttributeError(f"No {attr_name} found in _dynamic_imports for module name -> {__name__}")
+ try:
+ module = import_module(module_name, __package__)
+ if module_name == f".{attr_name}":
+ return module
+ else:
+ return getattr(module, attr_name)
+ except ImportError as e:
+ raise ImportError(f"Failed to import {attr_name} from {module_name}: {e}") from e
+ except AttributeError as e:
+ raise AttributeError(f"Failed to get {attr_name} from {module_name}: {e}") from e
+
+
+def __dir__():
+ lazy_attrs = list(_dynamic_imports.keys())
+ return sorted(lazy_attrs)
+
+
+__all__ = ["ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort"]
diff --git a/sdks/python/experiments/client.py b/sdks/python/experiments/client.py
new file mode 100644
index 0000000000..86893bba78
--- /dev/null
+++ b/sdks/python/experiments/client.py
@@ -0,0 +1,802 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.request_options import RequestOptions
+from ..types.experiment_filter import ExperimentFilter
+from ..types.experiment_response import ExperimentResponse
+from ..types.experiment_responses_page import ExperimentResponsesPage
+from ..types.experiment_session_filter import ExperimentSessionFilter
+from ..types.experiment_session_responses_page import ExperimentSessionResponsesPage
+from .raw_client import AsyncRawExperimentsClient, RawExperimentsClient
+from .types.list_experiments_apis_intake_v2workspaces_workspace_experiments_get_request_sort import (
+ ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort,
+)
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class ExperimentsClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._raw_client = RawExperimentsClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> RawExperimentsClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ RawExperimentsClient
+ """
+ return self._raw_client
+
+ def list_experiments(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort] = None,
+ filter: typing.Optional[ExperimentFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ExperimentResponsesPage:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort]
+ Sort field; prefix with '-' for descending.
+
+ filter : typing.Optional[ExperimentFilter]
+ Filter experiments by name, experiment_group_id, agent_name, agent_version, dataset_name, dataset_version, created_by, created_at, or updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ExperimentResponsesPage
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.experiments.list_experiments(
+ workspace="workspace",
+ )
+ """
+ _response = self._raw_client.list_experiments(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ def create_experiment(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ agent_name: str,
+ agent_version: str,
+ dataset_name: str,
+ experiment_group_id: typing.Optional[str] = OMIT,
+ dataset_version: typing.Optional[str] = OMIT,
+ source_link: typing.Optional[str] = OMIT,
+ metadata: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ summary: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ExperimentResponse:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Producer-supplied, workspace-unique experiment id.
+
+ agent_name : str
+ Name of the agent under test.
+
+ agent_version : str
+ Version of the agent under test.
+
+ dataset_name : str
+ Producer-supplied dataset name.
+
+ experiment_group_id : typing.Optional[str]
+ Entity id of the owning ExperimentGroup; optional. Soft reference, not validated.
+
+ dataset_version : typing.Optional[str]
+ Producer-supplied dataset version.
+
+ source_link : typing.Optional[str]
+ Optional URL for the source experiment.
+
+ metadata : typing.Optional[typing.Dict[str, typing.Any]]
+ Free-form producer metadata.
+
+ description : typing.Optional[str]
+ Human-readable description.
+
+ summary : typing.Optional[str]
+ Human-authored summary of results.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ExperimentResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.experiments.create_experiment(
+ workspace="workspace",
+ name="name",
+ agent_name="agent_name",
+ agent_version="agent_version",
+ dataset_name="dataset_name",
+ )
+ """
+ _response = self._raw_client.create_experiment(
+ workspace,
+ name=name,
+ agent_name=agent_name,
+ agent_version=agent_version,
+ dataset_name=dataset_name,
+ experiment_group_id=experiment_group_id,
+ dataset_version=dataset_version,
+ source_link=source_link,
+ metadata=metadata,
+ description=description,
+ summary=summary,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def get_experiment(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> ExperimentResponse:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ExperimentResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.experiments.get_experiment(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.get_experiment(workspace, name, request_options=request_options)
+ return _response.data
+
+ def update_experiment(
+ self,
+ workspace: str,
+ name_: str,
+ *,
+ name: str,
+ agent_name: str,
+ agent_version: str,
+ dataset_name: str,
+ experiment_group_id: typing.Optional[str] = OMIT,
+ dataset_version: typing.Optional[str] = OMIT,
+ source_link: typing.Optional[str] = OMIT,
+ metadata: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ summary: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ExperimentResponse:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name_ : str
+
+ name : str
+ Producer-supplied, workspace-unique experiment id.
+
+ agent_name : str
+ Name of the agent under test.
+
+ agent_version : str
+ Version of the agent under test.
+
+ dataset_name : str
+ Producer-supplied dataset name.
+
+ experiment_group_id : typing.Optional[str]
+ Entity id of the owning ExperimentGroup; optional. Soft reference, not validated.
+
+ dataset_version : typing.Optional[str]
+ Producer-supplied dataset version.
+
+ source_link : typing.Optional[str]
+ Optional URL for the source experiment.
+
+ metadata : typing.Optional[typing.Dict[str, typing.Any]]
+ Free-form producer metadata.
+
+ description : typing.Optional[str]
+ Human-readable description.
+
+ summary : typing.Optional[str]
+ Human-authored summary of results.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ExperimentResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.experiments.update_experiment(
+ workspace="workspace",
+ name_="name",
+ name="name",
+ agent_name="agent_name",
+ agent_version="agent_version",
+ dataset_name="dataset_name",
+ )
+ """
+ _response = self._raw_client.update_experiment(
+ workspace,
+ name_,
+ name=name,
+ agent_name=agent_name,
+ agent_version=agent_version,
+ dataset_name=dataset_name,
+ experiment_group_id=experiment_group_id,
+ dataset_version=dataset_version,
+ source_link=source_link,
+ metadata=metadata,
+ description=description,
+ summary=summary,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def delete_experiment(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> None:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.experiments.delete_experiment(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.delete_experiment(workspace, name, request_options=request_options)
+ return _response.data
+
+ def list_experiment_sessions(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ filter: typing.Optional[ExperimentSessionFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ExperimentSessionResponsesPage:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ filter : typing.Optional[ExperimentSessionFilter]
+ Filter sessions by test_case_id and status.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ExperimentSessionResponsesPage
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.experiments.list_experiment_sessions(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.list_experiment_sessions(
+ workspace, name, page=page, page_size=page_size, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+
+class AsyncExperimentsClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._raw_client = AsyncRawExperimentsClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> AsyncRawExperimentsClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ AsyncRawExperimentsClient
+ """
+ return self._raw_client
+
+ async def list_experiments(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort] = None,
+ filter: typing.Optional[ExperimentFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ExperimentResponsesPage:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort]
+ Sort field; prefix with '-' for descending.
+
+ filter : typing.Optional[ExperimentFilter]
+ Filter experiments by name, experiment_group_id, agent_name, agent_version, dataset_name, dataset_version, created_by, created_at, or updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ExperimentResponsesPage
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.experiments.list_experiments(
+ workspace="workspace",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_experiments(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ async def create_experiment(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ agent_name: str,
+ agent_version: str,
+ dataset_name: str,
+ experiment_group_id: typing.Optional[str] = OMIT,
+ dataset_version: typing.Optional[str] = OMIT,
+ source_link: typing.Optional[str] = OMIT,
+ metadata: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ summary: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ExperimentResponse:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Producer-supplied, workspace-unique experiment id.
+
+ agent_name : str
+ Name of the agent under test.
+
+ agent_version : str
+ Version of the agent under test.
+
+ dataset_name : str
+ Producer-supplied dataset name.
+
+ experiment_group_id : typing.Optional[str]
+ Entity id of the owning ExperimentGroup; optional. Soft reference, not validated.
+
+ dataset_version : typing.Optional[str]
+ Producer-supplied dataset version.
+
+ source_link : typing.Optional[str]
+ Optional URL for the source experiment.
+
+ metadata : typing.Optional[typing.Dict[str, typing.Any]]
+ Free-form producer metadata.
+
+ description : typing.Optional[str]
+ Human-readable description.
+
+ summary : typing.Optional[str]
+ Human-authored summary of results.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ExperimentResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.experiments.create_experiment(
+ workspace="workspace",
+ name="name",
+ agent_name="agent_name",
+ agent_version="agent_version",
+ dataset_name="dataset_name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.create_experiment(
+ workspace,
+ name=name,
+ agent_name=agent_name,
+ agent_version=agent_version,
+ dataset_name=dataset_name,
+ experiment_group_id=experiment_group_id,
+ dataset_version=dataset_version,
+ source_link=source_link,
+ metadata=metadata,
+ description=description,
+ summary=summary,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def get_experiment(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> ExperimentResponse:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ExperimentResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.experiments.get_experiment(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_experiment(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def update_experiment(
+ self,
+ workspace: str,
+ name_: str,
+ *,
+ name: str,
+ agent_name: str,
+ agent_version: str,
+ dataset_name: str,
+ experiment_group_id: typing.Optional[str] = OMIT,
+ dataset_version: typing.Optional[str] = OMIT,
+ source_link: typing.Optional[str] = OMIT,
+ metadata: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ summary: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ExperimentResponse:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name_ : str
+
+ name : str
+ Producer-supplied, workspace-unique experiment id.
+
+ agent_name : str
+ Name of the agent under test.
+
+ agent_version : str
+ Version of the agent under test.
+
+ dataset_name : str
+ Producer-supplied dataset name.
+
+ experiment_group_id : typing.Optional[str]
+ Entity id of the owning ExperimentGroup; optional. Soft reference, not validated.
+
+ dataset_version : typing.Optional[str]
+ Producer-supplied dataset version.
+
+ source_link : typing.Optional[str]
+ Optional URL for the source experiment.
+
+ metadata : typing.Optional[typing.Dict[str, typing.Any]]
+ Free-form producer metadata.
+
+ description : typing.Optional[str]
+ Human-readable description.
+
+ summary : typing.Optional[str]
+ Human-authored summary of results.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ExperimentResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.experiments.update_experiment(
+ workspace="workspace",
+ name_="name",
+ name="name",
+ agent_name="agent_name",
+ agent_version="agent_version",
+ dataset_name="dataset_name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.update_experiment(
+ workspace,
+ name_,
+ name=name,
+ agent_name=agent_name,
+ agent_version=agent_version,
+ dataset_name=dataset_name,
+ experiment_group_id=experiment_group_id,
+ dataset_version=dataset_version,
+ source_link=source_link,
+ metadata=metadata,
+ description=description,
+ summary=summary,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def delete_experiment(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> None:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.experiments.delete_experiment(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.delete_experiment(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def list_experiment_sessions(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ filter: typing.Optional[ExperimentSessionFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ExperimentSessionResponsesPage:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ filter : typing.Optional[ExperimentSessionFilter]
+ Filter sessions by test_case_id and status.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ExperimentSessionResponsesPage
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.experiments.list_experiment_sessions(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_experiment_sessions(
+ workspace, name, page=page, page_size=page_size, filter=filter, request_options=request_options
+ )
+ return _response.data
diff --git a/sdks/python/experiments/raw_client.py b/sdks/python/experiments/raw_client.py
new file mode 100644
index 0000000000..2be5e9fa37
--- /dev/null
+++ b/sdks/python/experiments/raw_client.py
@@ -0,0 +1,1143 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+from json.decoder import JSONDecodeError
+
+from ..core.api_error import ApiError
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.http_response import AsyncHttpResponse, HttpResponse
+from ..core.jsonable_encoder import encode_path_param
+from ..core.parse_error import ParsingError
+from ..core.pydantic_utilities import parse_obj_as
+from ..core.request_options import RequestOptions
+from ..core.serialization import convert_and_respect_annotation_metadata
+from ..errors.conflict_error import ConflictError
+from ..errors.not_found_error import NotFoundError
+from ..errors.service_unavailable_error import ServiceUnavailableError
+from ..errors.unprocessable_entity_error import UnprocessableEntityError
+from ..types.experiment_filter import ExperimentFilter
+from ..types.experiment_response import ExperimentResponse
+from ..types.experiment_responses_page import ExperimentResponsesPage
+from ..types.experiment_session_filter import ExperimentSessionFilter
+from ..types.experiment_session_responses_page import ExperimentSessionResponsesPage
+from .types.list_experiments_apis_intake_v2workspaces_workspace_experiments_get_request_sort import (
+ ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort,
+)
+from pydantic import ValidationError
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class RawExperimentsClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ def list_experiments(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort] = None,
+ filter: typing.Optional[ExperimentFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ExperimentResponsesPage]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort]
+ Sort field; prefix with '-' for descending.
+
+ filter : typing.Optional[ExperimentFilter]
+ Filter experiments by name, experiment_group_id, agent_name, agent_version, dataset_name, dataset_version, created_by, created_at, or updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ExperimentResponsesPage]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiments",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=ExperimentFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ExperimentResponsesPage,
+ parse_obj_as(
+ type_=ExperimentResponsesPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def create_experiment(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ agent_name: str,
+ agent_version: str,
+ dataset_name: str,
+ experiment_group_id: typing.Optional[str] = OMIT,
+ dataset_version: typing.Optional[str] = OMIT,
+ source_link: typing.Optional[str] = OMIT,
+ metadata: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ summary: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ExperimentResponse]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Producer-supplied, workspace-unique experiment id.
+
+ agent_name : str
+ Name of the agent under test.
+
+ agent_version : str
+ Version of the agent under test.
+
+ dataset_name : str
+ Producer-supplied dataset name.
+
+ experiment_group_id : typing.Optional[str]
+ Entity id of the owning ExperimentGroup; optional. Soft reference, not validated.
+
+ dataset_version : typing.Optional[str]
+ Producer-supplied dataset version.
+
+ source_link : typing.Optional[str]
+ Optional URL for the source experiment.
+
+ metadata : typing.Optional[typing.Dict[str, typing.Any]]
+ Free-form producer metadata.
+
+ description : typing.Optional[str]
+ Human-readable description.
+
+ summary : typing.Optional[str]
+ Human-authored summary of results.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ExperimentResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiments",
+ method="POST",
+ json={
+ "name": name,
+ "experiment_group_id": experiment_group_id,
+ "agent_name": agent_name,
+ "agent_version": agent_version,
+ "dataset_name": dataset_name,
+ "dataset_version": dataset_version,
+ "source_link": source_link,
+ "metadata": metadata,
+ "description": description,
+ "summary": summary,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ExperimentResponse,
+ parse_obj_as(
+ type_=ExperimentResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 409:
+ raise ConflictError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_experiment(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[ExperimentResponse]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ExperimentResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiments/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ExperimentResponse,
+ parse_obj_as(
+ type_=ExperimentResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def update_experiment(
+ self,
+ workspace: str,
+ name_: str,
+ *,
+ name: str,
+ agent_name: str,
+ agent_version: str,
+ dataset_name: str,
+ experiment_group_id: typing.Optional[str] = OMIT,
+ dataset_version: typing.Optional[str] = OMIT,
+ source_link: typing.Optional[str] = OMIT,
+ metadata: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ summary: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ExperimentResponse]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name_ : str
+
+ name : str
+ Producer-supplied, workspace-unique experiment id.
+
+ agent_name : str
+ Name of the agent under test.
+
+ agent_version : str
+ Version of the agent under test.
+
+ dataset_name : str
+ Producer-supplied dataset name.
+
+ experiment_group_id : typing.Optional[str]
+ Entity id of the owning ExperimentGroup; optional. Soft reference, not validated.
+
+ dataset_version : typing.Optional[str]
+ Producer-supplied dataset version.
+
+ source_link : typing.Optional[str]
+ Optional URL for the source experiment.
+
+ metadata : typing.Optional[typing.Dict[str, typing.Any]]
+ Free-form producer metadata.
+
+ description : typing.Optional[str]
+ Human-readable description.
+
+ summary : typing.Optional[str]
+ Human-authored summary of results.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ExperimentResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiments/{encode_path_param(name_)}",
+ method="PUT",
+ json={
+ "name": name,
+ "experiment_group_id": experiment_group_id,
+ "agent_name": agent_name,
+ "agent_version": agent_version,
+ "dataset_name": dataset_name,
+ "dataset_version": dataset_version,
+ "source_link": source_link,
+ "metadata": metadata,
+ "description": description,
+ "summary": summary,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ExperimentResponse,
+ parse_obj_as(
+ type_=ExperimentResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 409:
+ raise ConflictError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def delete_experiment(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[None]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[None]
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiments/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return HttpResponse(response=_response, data=None)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def list_experiment_sessions(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ filter: typing.Optional[ExperimentSessionFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ExperimentSessionResponsesPage]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ filter : typing.Optional[ExperimentSessionFilter]
+ Filter sessions by test_case_id and status.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ExperimentSessionResponsesPage]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiments/{encode_path_param(name)}/sessions",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=ExperimentSessionFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ExperimentSessionResponsesPage,
+ parse_obj_as(
+ type_=ExperimentSessionResponsesPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 503:
+ raise ServiceUnavailableError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+
+class AsyncRawExperimentsClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ async def list_experiments(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort] = None,
+ filter: typing.Optional[ExperimentFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ExperimentResponsesPage]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort]
+ Sort field; prefix with '-' for descending.
+
+ filter : typing.Optional[ExperimentFilter]
+ Filter experiments by name, experiment_group_id, agent_name, agent_version, dataset_name, dataset_version, created_by, created_at, or updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ExperimentResponsesPage]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiments",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=ExperimentFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ExperimentResponsesPage,
+ parse_obj_as(
+ type_=ExperimentResponsesPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def create_experiment(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ agent_name: str,
+ agent_version: str,
+ dataset_name: str,
+ experiment_group_id: typing.Optional[str] = OMIT,
+ dataset_version: typing.Optional[str] = OMIT,
+ source_link: typing.Optional[str] = OMIT,
+ metadata: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ summary: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ExperimentResponse]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Producer-supplied, workspace-unique experiment id.
+
+ agent_name : str
+ Name of the agent under test.
+
+ agent_version : str
+ Version of the agent under test.
+
+ dataset_name : str
+ Producer-supplied dataset name.
+
+ experiment_group_id : typing.Optional[str]
+ Entity id of the owning ExperimentGroup; optional. Soft reference, not validated.
+
+ dataset_version : typing.Optional[str]
+ Producer-supplied dataset version.
+
+ source_link : typing.Optional[str]
+ Optional URL for the source experiment.
+
+ metadata : typing.Optional[typing.Dict[str, typing.Any]]
+ Free-form producer metadata.
+
+ description : typing.Optional[str]
+ Human-readable description.
+
+ summary : typing.Optional[str]
+ Human-authored summary of results.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ExperimentResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiments",
+ method="POST",
+ json={
+ "name": name,
+ "experiment_group_id": experiment_group_id,
+ "agent_name": agent_name,
+ "agent_version": agent_version,
+ "dataset_name": dataset_name,
+ "dataset_version": dataset_version,
+ "source_link": source_link,
+ "metadata": metadata,
+ "description": description,
+ "summary": summary,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ExperimentResponse,
+ parse_obj_as(
+ type_=ExperimentResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 409:
+ raise ConflictError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_experiment(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[ExperimentResponse]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ExperimentResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiments/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ExperimentResponse,
+ parse_obj_as(
+ type_=ExperimentResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def update_experiment(
+ self,
+ workspace: str,
+ name_: str,
+ *,
+ name: str,
+ agent_name: str,
+ agent_version: str,
+ dataset_name: str,
+ experiment_group_id: typing.Optional[str] = OMIT,
+ dataset_version: typing.Optional[str] = OMIT,
+ source_link: typing.Optional[str] = OMIT,
+ metadata: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ summary: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ExperimentResponse]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name_ : str
+
+ name : str
+ Producer-supplied, workspace-unique experiment id.
+
+ agent_name : str
+ Name of the agent under test.
+
+ agent_version : str
+ Version of the agent under test.
+
+ dataset_name : str
+ Producer-supplied dataset name.
+
+ experiment_group_id : typing.Optional[str]
+ Entity id of the owning ExperimentGroup; optional. Soft reference, not validated.
+
+ dataset_version : typing.Optional[str]
+ Producer-supplied dataset version.
+
+ source_link : typing.Optional[str]
+ Optional URL for the source experiment.
+
+ metadata : typing.Optional[typing.Dict[str, typing.Any]]
+ Free-form producer metadata.
+
+ description : typing.Optional[str]
+ Human-readable description.
+
+ summary : typing.Optional[str]
+ Human-authored summary of results.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ExperimentResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiments/{encode_path_param(name_)}",
+ method="PUT",
+ json={
+ "name": name,
+ "experiment_group_id": experiment_group_id,
+ "agent_name": agent_name,
+ "agent_version": agent_version,
+ "dataset_name": dataset_name,
+ "dataset_version": dataset_version,
+ "source_link": source_link,
+ "metadata": metadata,
+ "description": description,
+ "summary": summary,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ExperimentResponse,
+ parse_obj_as(
+ type_=ExperimentResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 409:
+ raise ConflictError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def delete_experiment(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[None]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[None]
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiments/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return AsyncHttpResponse(response=_response, data=None)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def list_experiment_sessions(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ filter: typing.Optional[ExperimentSessionFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ExperimentSessionResponsesPage]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ filter : typing.Optional[ExperimentSessionFilter]
+ Filter sessions by test_case_id and status.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ExperimentSessionResponsesPage]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/experiments/{encode_path_param(name)}/sessions",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=ExperimentSessionFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ExperimentSessionResponsesPage,
+ parse_obj_as(
+ type_=ExperimentSessionResponsesPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 503:
+ raise ServiceUnavailableError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
diff --git a/sdks/python/experiments/types/__init__.py b/sdks/python/experiments/types/__init__.py
new file mode 100644
index 0000000000..b649223ace
--- /dev/null
+++ b/sdks/python/experiments/types/__init__.py
@@ -0,0 +1,38 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
+import typing
+from importlib import import_module
+
+if typing.TYPE_CHECKING:
+ from .list_experiments_apis_intake_v2workspaces_workspace_experiments_get_request_sort import (
+ ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort,
+ )
+_dynamic_imports: typing.Dict[str, str] = {
+ "ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort": ".list_experiments_apis_intake_v2workspaces_workspace_experiments_get_request_sort"
+}
+
+
+def __getattr__(attr_name: str) -> typing.Any:
+ module_name = _dynamic_imports.get(attr_name)
+ if module_name is None:
+ raise AttributeError(f"No {attr_name} found in _dynamic_imports for module name -> {__name__}")
+ try:
+ module = import_module(module_name, __package__)
+ if module_name == f".{attr_name}":
+ return module
+ else:
+ return getattr(module, attr_name)
+ except ImportError as e:
+ raise ImportError(f"Failed to import {attr_name} from {module_name}: {e}") from e
+ except AttributeError as e:
+ raise AttributeError(f"Failed to get {attr_name} from {module_name}: {e}") from e
+
+
+def __dir__():
+ lazy_attrs = list(_dynamic_imports.keys())
+ return sorted(lazy_attrs)
+
+
+__all__ = ["ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort"]
diff --git a/sdks/python/experiments/types/list_experiments_apis_intake_v2workspaces_workspace_experiments_get_request_sort.py b/sdks/python/experiments/types/list_experiments_apis_intake_v2workspaces_workspace_experiments_get_request_sort.py
new file mode 100644
index 0000000000..d5b7ebd098
--- /dev/null
+++ b/sdks/python/experiments/types/list_experiments_apis_intake_v2workspaces_workspace_experiments_get_request_sort.py
@@ -0,0 +1,7 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort = typing.Union[
+ typing.Literal["-created_at", "created_at", "-updated_at", "updated_at", "-name"], typing.Any
+]
diff --git a/sdks/python/files/__init__.py b/sdks/python/files/__init__.py
new file mode 100644
index 0000000000..3cde42948a
--- /dev/null
+++ b/sdks/python/files/__init__.py
@@ -0,0 +1,52 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
+import typing
+from importlib import import_module
+
+if typing.TYPE_CHECKING:
+ from .types import (
+ CreateFilesetRequestStorage,
+ CreateFilesetRequestStorage_Huggingface,
+ CreateFilesetRequestStorage_Local,
+ CreateFilesetRequestStorage_Ngc,
+ CreateFilesetRequestStorage_S3,
+ )
+_dynamic_imports: typing.Dict[str, str] = {
+ "CreateFilesetRequestStorage": ".types",
+ "CreateFilesetRequestStorage_Huggingface": ".types",
+ "CreateFilesetRequestStorage_Local": ".types",
+ "CreateFilesetRequestStorage_Ngc": ".types",
+ "CreateFilesetRequestStorage_S3": ".types",
+}
+
+
+def __getattr__(attr_name: str) -> typing.Any:
+ module_name = _dynamic_imports.get(attr_name)
+ if module_name is None:
+ raise AttributeError(f"No {attr_name} found in _dynamic_imports for module name -> {__name__}")
+ try:
+ module = import_module(module_name, __package__)
+ if module_name == f".{attr_name}":
+ return module
+ else:
+ return getattr(module, attr_name)
+ except ImportError as e:
+ raise ImportError(f"Failed to import {attr_name} from {module_name}: {e}") from e
+ except AttributeError as e:
+ raise AttributeError(f"Failed to get {attr_name} from {module_name}: {e}") from e
+
+
+def __dir__():
+ lazy_attrs = list(_dynamic_imports.keys())
+ return sorted(lazy_attrs)
+
+
+__all__ = [
+ "CreateFilesetRequestStorage",
+ "CreateFilesetRequestStorage_Huggingface",
+ "CreateFilesetRequestStorage_Local",
+ "CreateFilesetRequestStorage_Ngc",
+ "CreateFilesetRequestStorage_S3",
+]
diff --git a/sdks/python/files/client.py b/sdks/python/files/client.py
new file mode 100644
index 0000000000..58678b21d1
--- /dev/null
+++ b/sdks/python/files/client.py
@@ -0,0 +1,1136 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.request_options import RequestOptions
+from ..types.fileset_file_output import FilesetFileOutput
+from ..types.fileset_filter import FilesetFilter
+from ..types.fileset_metadata_input import FilesetMetadataInput
+from ..types.fileset_output import FilesetOutput
+from ..types.fileset_outputs_page import FilesetOutputsPage
+from ..types.fileset_purpose import FilesetPurpose
+from ..types.generic_sort_field import GenericSortField
+from ..types.list_fileset_files_response import ListFilesetFilesResponse
+from .raw_client import AsyncRawFilesClient, RawFilesClient
+from .types.create_fileset_request_storage import CreateFilesetRequestStorage
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class FilesClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._raw_client = RawFilesClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> RawFilesClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ RawFilesClient
+ """
+ return self._raw_client
+
+ def list_filesets(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[GenericSortField] = None,
+ filter: typing.Optional[FilesetFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> FilesetOutputsPage:
+ """
+ List Filesets endpoint with filtering and pagination.
+
+ Supports filtering by name, description, purpose, storage_type, created_at, and updated_at via query parameters.
+ Returns paginated results with sorting options.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[GenericSortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[FilesetFilter]
+ Filter filesets by name, description, purpose, storage_type, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ FilesetOutputsPage
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.files.list_filesets(
+ workspace="workspace",
+ )
+ """
+ _response = self._raw_client.list_filesets(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ def create_fileset(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ description: typing.Optional[str] = OMIT,
+ project: typing.Optional[str] = OMIT,
+ storage: typing.Optional[CreateFilesetRequestStorage] = OMIT,
+ purpose: typing.Optional[FilesetPurpose] = OMIT,
+ metadata: typing.Optional[FilesetMetadataInput] = OMIT,
+ custom_fields: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ cache: typing.Optional[bool] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> FilesetOutput:
+ """
+ Create a new fileset.
+
+ If no storage configuration is provided, the default storage backend will be used.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ The name of the fileset. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ description : typing.Optional[str]
+ The description of the fileset.
+
+ project : typing.Optional[str]
+ The name of the project associated with this fileset.
+
+ storage : typing.Optional[CreateFilesetRequestStorage]
+ The storage configuration for the fileset. If not provided, uses default storage.
+
+ purpose : typing.Optional[FilesetPurpose]
+ The purpose of the fileset.
+
+ metadata : typing.Optional[FilesetMetadataInput]
+ Purpose-specific metadata. Use the purpose as the key (e.g., {dataset: {...}}).
+
+ custom_fields : typing.Optional[typing.Dict[str, typing.Any]]
+ Custom fields for the fileset.
+
+ cache : typing.Optional[bool]
+ Cache all files after creation. Only applies to external storage.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ FilesetOutput
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.files.create_fileset(
+ workspace="workspace",
+ name="training-data-v1",
+ )
+ """
+ _response = self._raw_client.create_fileset(
+ workspace,
+ name=name,
+ description=description,
+ project=project,
+ storage=storage,
+ purpose=purpose,
+ metadata=metadata,
+ custom_fields=custom_fields,
+ cache=cache,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def retrieve_fileset(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> FilesetOutput:
+ """
+ Get Fileset by Workspace and Name.
+
+ Returns the details of a specific fileset identified by its workspace and name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ FilesetOutput
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.files.retrieve_fileset(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.retrieve_fileset(workspace, name, request_options=request_options)
+ return _response.data
+
+ def delete_fileset(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> FilesetOutput:
+ """
+ Delete Fileset.
+
+ Permanently deletes a fileset from the platform.
+ Returns metadata about the deleted fileset.
+ For local storage backends, this also deletes the underlying files.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ FilesetOutput
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.files.delete_fileset(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.delete_fileset(workspace, name, request_options=request_options)
+ return _response.data
+
+ def update_fileset_metadata(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ project: typing.Optional[str] = OMIT,
+ purpose: typing.Optional[FilesetPurpose] = OMIT,
+ metadata: typing.Optional[FilesetMetadataInput] = OMIT,
+ custom_fields: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> FilesetOutput:
+ """
+ Update Fileset Metadata.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ description : typing.Optional[str]
+ The description of the fileset.
+
+ project : typing.Optional[str]
+ The name of the project associated with this fileset.
+
+ purpose : typing.Optional[FilesetPurpose]
+ The purpose of the fileset.
+
+ metadata : typing.Optional[FilesetMetadataInput]
+ Purpose-specific metadata. Use the purpose as the key (e.g., {dataset: {...}}).
+
+ custom_fields : typing.Optional[typing.Dict[str, typing.Any]]
+ Custom fields for the fileset.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ FilesetOutput
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.files.update_fileset_metadata(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.update_fileset_metadata(
+ workspace,
+ name,
+ description=description,
+ project=project,
+ purpose=purpose,
+ metadata=metadata,
+ custom_fields=custom_fields,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def download_file_apis_files_v2workspaces_workspace_filesets_name_path_get(
+ self, workspace: str, name: str, path: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Iterator[bytes]:
+ """
+ Download file content from a fileset.
+
+ Supports HTTP Range requests for partial content retrieval (status 206).
+ Returns the full file content (status 200) if no Range header is provided.
+ For external resources (HuggingFace, NGC), content is cached locally on first access.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ path : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration. You can pass in configuration such as `chunk_size`, and more to customize the request and response.
+
+ Returns
+ -------
+ typing.Iterator[bytes]
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.files.download_file_apis_files_v2workspaces_workspace_filesets_name_path_get(
+ workspace="workspace",
+ name="name",
+ path="path",
+ )
+ """
+ with self._raw_client.download_file_apis_files_v2workspaces_workspace_filesets_name_path_get(
+ workspace, name, path, request_options=request_options
+ ) as r:
+ yield from r.data
+
+ def upload_file_apis_files_v2workspaces_workspace_filesets_name_path_put(
+ self,
+ workspace: str,
+ name: str,
+ path: str,
+ *,
+ request: typing.Union[bytes, typing.Iterator[bytes], typing.AsyncIterator[bytes]],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> FilesetFileOutput:
+ """
+ Upload file content to a fileset.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ path : str
+
+ request : typing.Union[bytes, typing.Iterator[bytes], typing.AsyncIterator[bytes]]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ FilesetFileOutput
+ Successful Response
+ """
+ _response = self._raw_client.upload_file_apis_files_v2workspaces_workspace_filesets_name_path_put(
+ workspace, name, path, request=request, request_options=request_options
+ )
+ return _response.data
+
+ def delete_file_apis_files_v2workspaces_workspace_filesets_name_path_delete(
+ self, workspace: str, name: str, path: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> FilesetFileOutput:
+ """
+ Delete a specific file from a fileset.
+
+ Permanently deletes the file from the storage backend.
+ Returns metadata about the deleted file.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ path : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ FilesetFileOutput
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.files.delete_file_apis_files_v2workspaces_workspace_filesets_name_path_delete(
+ workspace="workspace",
+ name="name",
+ path="path",
+ )
+ """
+ _response = self._raw_client.delete_file_apis_files_v2workspaces_workspace_filesets_name_path_delete(
+ workspace, name, path, request_options=request_options
+ )
+ return _response.data
+
+ def head_file_apis_files_v2workspaces_workspace_filesets_name_path_head(
+ self, workspace: str, name: str, path: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Dict[str, str]:
+ """
+ Get file metadata without downloading content.
+
+ HEAD requests are often used before Range GETs to ensure the server
+ supports partial downloads (e.g., DuckDB's httpfs).
+ Returns Accept-Ranges, Content-Length, and Content-Type headers.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ path : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, str]
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.files.head_file_apis_files_v2workspaces_workspace_filesets_name_path_head(
+ workspace="workspace",
+ name="name",
+ path="path",
+ )
+ """
+ _response = self._raw_client.head_file_apis_files_v2workspaces_workspace_filesets_name_path_head(
+ workspace, name, path, request_options=request_options
+ )
+ return _response.headers
+
+ def list_fileset_files(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ path: typing.Optional[str] = None,
+ include_cache_status: typing.Optional[bool] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ListFilesetFilesResponse:
+ """
+ List Files in Fileset.
+
+ Returns a list of files stored in the specified fileset.
+ Optionally filter by path prefix to list files under a specific directory.
+
+ Each file includes a cache_status field:
+ - "not_cacheable": File is on default storage, caching not applicable
+ - "cached": File exists in cache storage
+ - "caching": File is currently being downloaded and cached
+ - "not_cached": File not in cache, will be cached on next download
+ - null: External storage, but cache status not checked (use include_cache_status=true)
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ path : typing.Optional[str]
+ Filter files by path prefix
+
+ include_cache_status : typing.Optional[bool]
+ Check and return cache status for each file. When false, storage files return null for cache_status.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ListFilesetFilesResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.files.list_fileset_files(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.list_fileset_files(
+ workspace, name, path=path, include_cache_status=include_cache_status, request_options=request_options
+ )
+ return _response.data
+
+
+class AsyncFilesClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._raw_client = AsyncRawFilesClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> AsyncRawFilesClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ AsyncRawFilesClient
+ """
+ return self._raw_client
+
+ async def list_filesets(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[GenericSortField] = None,
+ filter: typing.Optional[FilesetFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> FilesetOutputsPage:
+ """
+ List Filesets endpoint with filtering and pagination.
+
+ Supports filtering by name, description, purpose, storage_type, created_at, and updated_at via query parameters.
+ Returns paginated results with sorting options.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[GenericSortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[FilesetFilter]
+ Filter filesets by name, description, purpose, storage_type, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ FilesetOutputsPage
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.files.list_filesets(
+ workspace="workspace",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_filesets(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ async def create_fileset(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ description: typing.Optional[str] = OMIT,
+ project: typing.Optional[str] = OMIT,
+ storage: typing.Optional[CreateFilesetRequestStorage] = OMIT,
+ purpose: typing.Optional[FilesetPurpose] = OMIT,
+ metadata: typing.Optional[FilesetMetadataInput] = OMIT,
+ custom_fields: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ cache: typing.Optional[bool] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> FilesetOutput:
+ """
+ Create a new fileset.
+
+ If no storage configuration is provided, the default storage backend will be used.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ The name of the fileset. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ description : typing.Optional[str]
+ The description of the fileset.
+
+ project : typing.Optional[str]
+ The name of the project associated with this fileset.
+
+ storage : typing.Optional[CreateFilesetRequestStorage]
+ The storage configuration for the fileset. If not provided, uses default storage.
+
+ purpose : typing.Optional[FilesetPurpose]
+ The purpose of the fileset.
+
+ metadata : typing.Optional[FilesetMetadataInput]
+ Purpose-specific metadata. Use the purpose as the key (e.g., {dataset: {...}}).
+
+ custom_fields : typing.Optional[typing.Dict[str, typing.Any]]
+ Custom fields for the fileset.
+
+ cache : typing.Optional[bool]
+ Cache all files after creation. Only applies to external storage.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ FilesetOutput
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.files.create_fileset(
+ workspace="workspace",
+ name="training-data-v1",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.create_fileset(
+ workspace,
+ name=name,
+ description=description,
+ project=project,
+ storage=storage,
+ purpose=purpose,
+ metadata=metadata,
+ custom_fields=custom_fields,
+ cache=cache,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def retrieve_fileset(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> FilesetOutput:
+ """
+ Get Fileset by Workspace and Name.
+
+ Returns the details of a specific fileset identified by its workspace and name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ FilesetOutput
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.files.retrieve_fileset(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.retrieve_fileset(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def delete_fileset(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> FilesetOutput:
+ """
+ Delete Fileset.
+
+ Permanently deletes a fileset from the platform.
+ Returns metadata about the deleted fileset.
+ For local storage backends, this also deletes the underlying files.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ FilesetOutput
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.files.delete_fileset(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.delete_fileset(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def update_fileset_metadata(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ project: typing.Optional[str] = OMIT,
+ purpose: typing.Optional[FilesetPurpose] = OMIT,
+ metadata: typing.Optional[FilesetMetadataInput] = OMIT,
+ custom_fields: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> FilesetOutput:
+ """
+ Update Fileset Metadata.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ description : typing.Optional[str]
+ The description of the fileset.
+
+ project : typing.Optional[str]
+ The name of the project associated with this fileset.
+
+ purpose : typing.Optional[FilesetPurpose]
+ The purpose of the fileset.
+
+ metadata : typing.Optional[FilesetMetadataInput]
+ Purpose-specific metadata. Use the purpose as the key (e.g., {dataset: {...}}).
+
+ custom_fields : typing.Optional[typing.Dict[str, typing.Any]]
+ Custom fields for the fileset.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ FilesetOutput
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.files.update_fileset_metadata(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.update_fileset_metadata(
+ workspace,
+ name,
+ description=description,
+ project=project,
+ purpose=purpose,
+ metadata=metadata,
+ custom_fields=custom_fields,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def download_file_apis_files_v2workspaces_workspace_filesets_name_path_get(
+ self, workspace: str, name: str, path: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.AsyncIterator[bytes]:
+ """
+ Download file content from a fileset.
+
+ Supports HTTP Range requests for partial content retrieval (status 206).
+ Returns the full file content (status 200) if no Range header is provided.
+ For external resources (HuggingFace, NGC), content is cached locally on first access.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ path : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration. You can pass in configuration such as `chunk_size`, and more to customize the request and response.
+
+ Returns
+ -------
+ typing.AsyncIterator[bytes]
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.files.download_file_apis_files_v2workspaces_workspace_filesets_name_path_get(
+ workspace="workspace",
+ name="name",
+ path="path",
+ )
+
+
+ asyncio.run(main())
+ """
+ async with self._raw_client.download_file_apis_files_v2workspaces_workspace_filesets_name_path_get(
+ workspace, name, path, request_options=request_options
+ ) as r:
+ async for _chunk in r.data:
+ yield _chunk
+
+ async def upload_file_apis_files_v2workspaces_workspace_filesets_name_path_put(
+ self,
+ workspace: str,
+ name: str,
+ path: str,
+ *,
+ request: typing.Union[bytes, typing.Iterator[bytes], typing.AsyncIterator[bytes]],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> FilesetFileOutput:
+ """
+ Upload file content to a fileset.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ path : str
+
+ request : typing.Union[bytes, typing.Iterator[bytes], typing.AsyncIterator[bytes]]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ FilesetFileOutput
+ Successful Response
+ """
+ _response = await self._raw_client.upload_file_apis_files_v2workspaces_workspace_filesets_name_path_put(
+ workspace, name, path, request=request, request_options=request_options
+ )
+ return _response.data
+
+ async def delete_file_apis_files_v2workspaces_workspace_filesets_name_path_delete(
+ self, workspace: str, name: str, path: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> FilesetFileOutput:
+ """
+ Delete a specific file from a fileset.
+
+ Permanently deletes the file from the storage backend.
+ Returns metadata about the deleted file.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ path : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ FilesetFileOutput
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.files.delete_file_apis_files_v2workspaces_workspace_filesets_name_path_delete(
+ workspace="workspace",
+ name="name",
+ path="path",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.delete_file_apis_files_v2workspaces_workspace_filesets_name_path_delete(
+ workspace, name, path, request_options=request_options
+ )
+ return _response.data
+
+ async def head_file_apis_files_v2workspaces_workspace_filesets_name_path_head(
+ self, workspace: str, name: str, path: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Dict[str, str]:
+ """
+ Get file metadata without downloading content.
+
+ HEAD requests are often used before Range GETs to ensure the server
+ supports partial downloads (e.g., DuckDB's httpfs).
+ Returns Accept-Ranges, Content-Length, and Content-Type headers.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ path : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, str]
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.files.head_file_apis_files_v2workspaces_workspace_filesets_name_path_head(
+ workspace="workspace",
+ name="name",
+ path="path",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.head_file_apis_files_v2workspaces_workspace_filesets_name_path_head(
+ workspace, name, path, request_options=request_options
+ )
+ return _response.headers
+
+ async def list_fileset_files(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ path: typing.Optional[str] = None,
+ include_cache_status: typing.Optional[bool] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ListFilesetFilesResponse:
+ """
+ List Files in Fileset.
+
+ Returns a list of files stored in the specified fileset.
+ Optionally filter by path prefix to list files under a specific directory.
+
+ Each file includes a cache_status field:
+ - "not_cacheable": File is on default storage, caching not applicable
+ - "cached": File exists in cache storage
+ - "caching": File is currently being downloaded and cached
+ - "not_cached": File not in cache, will be cached on next download
+ - null: External storage, but cache status not checked (use include_cache_status=true)
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ path : typing.Optional[str]
+ Filter files by path prefix
+
+ include_cache_status : typing.Optional[bool]
+ Check and return cache status for each file. When false, storage files return null for cache_status.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ListFilesetFilesResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.files.list_fileset_files(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_fileset_files(
+ workspace, name, path=path, include_cache_status=include_cache_status, request_options=request_options
+ )
+ return _response.data
diff --git a/sdks/python/files/raw_client.py b/sdks/python/files/raw_client.py
new file mode 100644
index 0000000000..ab702ccb4d
--- /dev/null
+++ b/sdks/python/files/raw_client.py
@@ -0,0 +1,1521 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import contextlib
+import typing
+from json.decoder import JSONDecodeError
+
+from ..core.api_error import ApiError
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.http_response import AsyncHttpResponse, HttpResponse
+from ..core.jsonable_encoder import encode_path_param
+from ..core.parse_error import ParsingError
+from ..core.pydantic_utilities import parse_obj_as
+from ..core.request_options import RequestOptions
+from ..core.serialization import convert_and_respect_annotation_metadata
+from ..errors.unprocessable_entity_error import UnprocessableEntityError
+from ..types.fileset_file_output import FilesetFileOutput
+from ..types.fileset_filter import FilesetFilter
+from ..types.fileset_metadata_input import FilesetMetadataInput
+from ..types.fileset_output import FilesetOutput
+from ..types.fileset_outputs_page import FilesetOutputsPage
+from ..types.fileset_purpose import FilesetPurpose
+from ..types.generic_sort_field import GenericSortField
+from ..types.list_fileset_files_response import ListFilesetFilesResponse
+from .types.create_fileset_request_storage import CreateFilesetRequestStorage
+from pydantic import ValidationError
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class RawFilesClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ def list_filesets(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[GenericSortField] = None,
+ filter: typing.Optional[FilesetFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[FilesetOutputsPage]:
+ """
+ List Filesets endpoint with filtering and pagination.
+
+ Supports filtering by name, description, purpose, storage_type, created_at, and updated_at via query parameters.
+ Returns paginated results with sorting options.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[GenericSortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[FilesetFilter]
+ Filter filesets by name, description, purpose, storage_type, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[FilesetOutputsPage]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=FilesetFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ FilesetOutputsPage,
+ parse_obj_as(
+ type_=FilesetOutputsPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def create_fileset(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ description: typing.Optional[str] = OMIT,
+ project: typing.Optional[str] = OMIT,
+ storage: typing.Optional[CreateFilesetRequestStorage] = OMIT,
+ purpose: typing.Optional[FilesetPurpose] = OMIT,
+ metadata: typing.Optional[FilesetMetadataInput] = OMIT,
+ custom_fields: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ cache: typing.Optional[bool] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[FilesetOutput]:
+ """
+ Create a new fileset.
+
+ If no storage configuration is provided, the default storage backend will be used.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ The name of the fileset. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ description : typing.Optional[str]
+ The description of the fileset.
+
+ project : typing.Optional[str]
+ The name of the project associated with this fileset.
+
+ storage : typing.Optional[CreateFilesetRequestStorage]
+ The storage configuration for the fileset. If not provided, uses default storage.
+
+ purpose : typing.Optional[FilesetPurpose]
+ The purpose of the fileset.
+
+ metadata : typing.Optional[FilesetMetadataInput]
+ Purpose-specific metadata. Use the purpose as the key (e.g., {dataset: {...}}).
+
+ custom_fields : typing.Optional[typing.Dict[str, typing.Any]]
+ Custom fields for the fileset.
+
+ cache : typing.Optional[bool]
+ Cache all files after creation. Only applies to external storage.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[FilesetOutput]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets",
+ method="POST",
+ json={
+ "name": name,
+ "description": description,
+ "project": project,
+ "storage": convert_and_respect_annotation_metadata(
+ object_=storage, annotation=CreateFilesetRequestStorage, direction="write"
+ ),
+ "purpose": purpose,
+ "metadata": convert_and_respect_annotation_metadata(
+ object_=metadata, annotation=FilesetMetadataInput, direction="write"
+ ),
+ "custom_fields": custom_fields,
+ "cache": cache,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ FilesetOutput,
+ parse_obj_as(
+ type_=FilesetOutput, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def retrieve_fileset(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[FilesetOutput]:
+ """
+ Get Fileset by Workspace and Name.
+
+ Returns the details of a specific fileset identified by its workspace and name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[FilesetOutput]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ FilesetOutput,
+ parse_obj_as(
+ type_=FilesetOutput, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def delete_fileset(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[FilesetOutput]:
+ """
+ Delete Fileset.
+
+ Permanently deletes a fileset from the platform.
+ Returns metadata about the deleted fileset.
+ For local storage backends, this also deletes the underlying files.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[FilesetOutput]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ FilesetOutput,
+ parse_obj_as(
+ type_=FilesetOutput, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def update_fileset_metadata(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ project: typing.Optional[str] = OMIT,
+ purpose: typing.Optional[FilesetPurpose] = OMIT,
+ metadata: typing.Optional[FilesetMetadataInput] = OMIT,
+ custom_fields: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[FilesetOutput]:
+ """
+ Update Fileset Metadata.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ description : typing.Optional[str]
+ The description of the fileset.
+
+ project : typing.Optional[str]
+ The name of the project associated with this fileset.
+
+ purpose : typing.Optional[FilesetPurpose]
+ The purpose of the fileset.
+
+ metadata : typing.Optional[FilesetMetadataInput]
+ Purpose-specific metadata. Use the purpose as the key (e.g., {dataset: {...}}).
+
+ custom_fields : typing.Optional[typing.Dict[str, typing.Any]]
+ Custom fields for the fileset.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[FilesetOutput]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets/{encode_path_param(name)}",
+ method="PATCH",
+ json={
+ "description": description,
+ "project": project,
+ "purpose": purpose,
+ "metadata": convert_and_respect_annotation_metadata(
+ object_=metadata, annotation=FilesetMetadataInput, direction="write"
+ ),
+ "custom_fields": custom_fields,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ FilesetOutput,
+ parse_obj_as(
+ type_=FilesetOutput, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ @contextlib.contextmanager
+ def download_file_apis_files_v2workspaces_workspace_filesets_name_path_get(
+ self, workspace: str, name: str, path: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Iterator[HttpResponse[typing.Iterator[bytes]]]:
+ """
+ Download file content from a fileset.
+
+ Supports HTTP Range requests for partial content retrieval (status 206).
+ Returns the full file content (status 200) if no Range header is provided.
+ For external resources (HuggingFace, NGC), content is cached locally on first access.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ path : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration. You can pass in configuration such as `chunk_size`, and more to customize the request and response.
+
+ Returns
+ -------
+ typing.Iterator[HttpResponse[typing.Iterator[bytes]]]
+ Successful Response
+ """
+ with self._client_wrapper.httpx_client.stream(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets/{encode_path_param(name)}/-/{encode_path_param(path)}",
+ method="GET",
+ request_options=request_options,
+ ) as _response:
+
+ def _stream() -> HttpResponse[typing.Iterator[bytes]]:
+ try:
+ if 200 <= _response.status_code < 300:
+ _chunk_size = request_options.get("chunk_size", None) if request_options is not None else None
+ return HttpResponse(
+ response=_response, data=(_chunk for _chunk in _response.iter_bytes(chunk_size=_chunk_size))
+ )
+ _response.read()
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.text
+ )
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code,
+ headers=dict(_response.headers),
+ body=_response.json(),
+ cause=e,
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ yield _stream()
+
+ def upload_file_apis_files_v2workspaces_workspace_filesets_name_path_put(
+ self,
+ workspace: str,
+ name: str,
+ path: str,
+ *,
+ request: typing.Union[bytes, typing.Iterator[bytes], typing.AsyncIterator[bytes]],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[FilesetFileOutput]:
+ """
+ Upload file content to a fileset.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ path : str
+
+ request : typing.Union[bytes, typing.Iterator[bytes], typing.AsyncIterator[bytes]]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[FilesetFileOutput]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets/{encode_path_param(name)}/-/{encode_path_param(path)}",
+ method="PUT",
+ content=request,
+ headers={
+ "content-type": "application/octet-stream",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ FilesetFileOutput,
+ parse_obj_as(
+ type_=FilesetFileOutput, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def delete_file_apis_files_v2workspaces_workspace_filesets_name_path_delete(
+ self, workspace: str, name: str, path: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[FilesetFileOutput]:
+ """
+ Delete a specific file from a fileset.
+
+ Permanently deletes the file from the storage backend.
+ Returns metadata about the deleted file.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ path : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[FilesetFileOutput]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets/{encode_path_param(name)}/-/{encode_path_param(path)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ FilesetFileOutput,
+ parse_obj_as(
+ type_=FilesetFileOutput, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def head_file_apis_files_v2workspaces_workspace_filesets_name_path_head(
+ self, workspace: str, name: str, path: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[typing.Any]:
+ """
+ Get file metadata without downloading content.
+
+ HEAD requests are often used before Range GETs to ensure the server
+ supports partial downloads (e.g., DuckDB's httpfs).
+ Returns Accept-Ranges, Content-Length, and Content-Type headers.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ path : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Any]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets/{encode_path_param(name)}/-/{encode_path_param(path)}",
+ method="HEAD",
+ request_options=request_options,
+ )
+ try:
+ if _response is None or not _response.text.strip():
+ return HttpResponse(response=_response, data=None)
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def list_fileset_files(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ path: typing.Optional[str] = None,
+ include_cache_status: typing.Optional[bool] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ListFilesetFilesResponse]:
+ """
+ List Files in Fileset.
+
+ Returns a list of files stored in the specified fileset.
+ Optionally filter by path prefix to list files under a specific directory.
+
+ Each file includes a cache_status field:
+ - "not_cacheable": File is on default storage, caching not applicable
+ - "cached": File exists in cache storage
+ - "caching": File is currently being downloaded and cached
+ - "not_cached": File not in cache, will be cached on next download
+ - null: External storage, but cache status not checked (use include_cache_status=true)
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ path : typing.Optional[str]
+ Filter files by path prefix
+
+ include_cache_status : typing.Optional[bool]
+ Check and return cache status for each file. When false, storage files return null for cache_status.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ListFilesetFilesResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets/{encode_path_param(name)}/files",
+ method="GET",
+ params={
+ "path": path,
+ "include_cache_status": include_cache_status,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ListFilesetFilesResponse,
+ parse_obj_as(
+ type_=ListFilesetFilesResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+
+class AsyncRawFilesClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ async def list_filesets(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[GenericSortField] = None,
+ filter: typing.Optional[FilesetFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[FilesetOutputsPage]:
+ """
+ List Filesets endpoint with filtering and pagination.
+
+ Supports filtering by name, description, purpose, storage_type, created_at, and updated_at via query parameters.
+ Returns paginated results with sorting options.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[GenericSortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[FilesetFilter]
+ Filter filesets by name, description, purpose, storage_type, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[FilesetOutputsPage]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=FilesetFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ FilesetOutputsPage,
+ parse_obj_as(
+ type_=FilesetOutputsPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def create_fileset(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ description: typing.Optional[str] = OMIT,
+ project: typing.Optional[str] = OMIT,
+ storage: typing.Optional[CreateFilesetRequestStorage] = OMIT,
+ purpose: typing.Optional[FilesetPurpose] = OMIT,
+ metadata: typing.Optional[FilesetMetadataInput] = OMIT,
+ custom_fields: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ cache: typing.Optional[bool] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[FilesetOutput]:
+ """
+ Create a new fileset.
+
+ If no storage configuration is provided, the default storage backend will be used.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ The name of the fileset. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ description : typing.Optional[str]
+ The description of the fileset.
+
+ project : typing.Optional[str]
+ The name of the project associated with this fileset.
+
+ storage : typing.Optional[CreateFilesetRequestStorage]
+ The storage configuration for the fileset. If not provided, uses default storage.
+
+ purpose : typing.Optional[FilesetPurpose]
+ The purpose of the fileset.
+
+ metadata : typing.Optional[FilesetMetadataInput]
+ Purpose-specific metadata. Use the purpose as the key (e.g., {dataset: {...}}).
+
+ custom_fields : typing.Optional[typing.Dict[str, typing.Any]]
+ Custom fields for the fileset.
+
+ cache : typing.Optional[bool]
+ Cache all files after creation. Only applies to external storage.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[FilesetOutput]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets",
+ method="POST",
+ json={
+ "name": name,
+ "description": description,
+ "project": project,
+ "storage": convert_and_respect_annotation_metadata(
+ object_=storage, annotation=CreateFilesetRequestStorage, direction="write"
+ ),
+ "purpose": purpose,
+ "metadata": convert_and_respect_annotation_metadata(
+ object_=metadata, annotation=FilesetMetadataInput, direction="write"
+ ),
+ "custom_fields": custom_fields,
+ "cache": cache,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ FilesetOutput,
+ parse_obj_as(
+ type_=FilesetOutput, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def retrieve_fileset(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[FilesetOutput]:
+ """
+ Get Fileset by Workspace and Name.
+
+ Returns the details of a specific fileset identified by its workspace and name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[FilesetOutput]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ FilesetOutput,
+ parse_obj_as(
+ type_=FilesetOutput, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def delete_fileset(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[FilesetOutput]:
+ """
+ Delete Fileset.
+
+ Permanently deletes a fileset from the platform.
+ Returns metadata about the deleted fileset.
+ For local storage backends, this also deletes the underlying files.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[FilesetOutput]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ FilesetOutput,
+ parse_obj_as(
+ type_=FilesetOutput, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def update_fileset_metadata(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ project: typing.Optional[str] = OMIT,
+ purpose: typing.Optional[FilesetPurpose] = OMIT,
+ metadata: typing.Optional[FilesetMetadataInput] = OMIT,
+ custom_fields: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[FilesetOutput]:
+ """
+ Update Fileset Metadata.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ description : typing.Optional[str]
+ The description of the fileset.
+
+ project : typing.Optional[str]
+ The name of the project associated with this fileset.
+
+ purpose : typing.Optional[FilesetPurpose]
+ The purpose of the fileset.
+
+ metadata : typing.Optional[FilesetMetadataInput]
+ Purpose-specific metadata. Use the purpose as the key (e.g., {dataset: {...}}).
+
+ custom_fields : typing.Optional[typing.Dict[str, typing.Any]]
+ Custom fields for the fileset.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[FilesetOutput]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets/{encode_path_param(name)}",
+ method="PATCH",
+ json={
+ "description": description,
+ "project": project,
+ "purpose": purpose,
+ "metadata": convert_and_respect_annotation_metadata(
+ object_=metadata, annotation=FilesetMetadataInput, direction="write"
+ ),
+ "custom_fields": custom_fields,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ FilesetOutput,
+ parse_obj_as(
+ type_=FilesetOutput, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ @contextlib.asynccontextmanager
+ async def download_file_apis_files_v2workspaces_workspace_filesets_name_path_get(
+ self, workspace: str, name: str, path: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.AsyncIterator[AsyncHttpResponse[typing.AsyncIterator[bytes]]]:
+ """
+ Download file content from a fileset.
+
+ Supports HTTP Range requests for partial content retrieval (status 206).
+ Returns the full file content (status 200) if no Range header is provided.
+ For external resources (HuggingFace, NGC), content is cached locally on first access.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ path : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration. You can pass in configuration such as `chunk_size`, and more to customize the request and response.
+
+ Returns
+ -------
+ typing.AsyncIterator[AsyncHttpResponse[typing.AsyncIterator[bytes]]]
+ Successful Response
+ """
+ async with self._client_wrapper.httpx_client.stream(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets/{encode_path_param(name)}/-/{encode_path_param(path)}",
+ method="GET",
+ request_options=request_options,
+ ) as _response:
+
+ async def _stream() -> AsyncHttpResponse[typing.AsyncIterator[bytes]]:
+ try:
+ if 200 <= _response.status_code < 300:
+ _chunk_size = request_options.get("chunk_size", None) if request_options is not None else None
+ return AsyncHttpResponse(
+ response=_response,
+ data=(_chunk async for _chunk in _response.aiter_bytes(chunk_size=_chunk_size)),
+ )
+ await _response.aread()
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.text
+ )
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code,
+ headers=dict(_response.headers),
+ body=_response.json(),
+ cause=e,
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ yield await _stream()
+
+ async def upload_file_apis_files_v2workspaces_workspace_filesets_name_path_put(
+ self,
+ workspace: str,
+ name: str,
+ path: str,
+ *,
+ request: typing.Union[bytes, typing.Iterator[bytes], typing.AsyncIterator[bytes]],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[FilesetFileOutput]:
+ """
+ Upload file content to a fileset.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ path : str
+
+ request : typing.Union[bytes, typing.Iterator[bytes], typing.AsyncIterator[bytes]]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[FilesetFileOutput]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets/{encode_path_param(name)}/-/{encode_path_param(path)}",
+ method="PUT",
+ content=request,
+ headers={
+ "content-type": "application/octet-stream",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ FilesetFileOutput,
+ parse_obj_as(
+ type_=FilesetFileOutput, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def delete_file_apis_files_v2workspaces_workspace_filesets_name_path_delete(
+ self, workspace: str, name: str, path: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[FilesetFileOutput]:
+ """
+ Delete a specific file from a fileset.
+
+ Permanently deletes the file from the storage backend.
+ Returns metadata about the deleted file.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ path : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[FilesetFileOutput]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets/{encode_path_param(name)}/-/{encode_path_param(path)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ FilesetFileOutput,
+ parse_obj_as(
+ type_=FilesetFileOutput, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def head_file_apis_files_v2workspaces_workspace_filesets_name_path_head(
+ self, workspace: str, name: str, path: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[typing.Any]:
+ """
+ Get file metadata without downloading content.
+
+ HEAD requests are often used before Range GETs to ensure the server
+ supports partial downloads (e.g., DuckDB's httpfs).
+ Returns Accept-Ranges, Content-Length, and Content-Type headers.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ path : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Any]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets/{encode_path_param(name)}/-/{encode_path_param(path)}",
+ method="HEAD",
+ request_options=request_options,
+ )
+ try:
+ if _response is None or not _response.text.strip():
+ return AsyncHttpResponse(response=_response, data=None)
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def list_fileset_files(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ path: typing.Optional[str] = None,
+ include_cache_status: typing.Optional[bool] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ListFilesetFilesResponse]:
+ """
+ List Files in Fileset.
+
+ Returns a list of files stored in the specified fileset.
+ Optionally filter by path prefix to list files under a specific directory.
+
+ Each file includes a cache_status field:
+ - "not_cacheable": File is on default storage, caching not applicable
+ - "cached": File exists in cache storage
+ - "caching": File is currently being downloaded and cached
+ - "not_cached": File not in cache, will be cached on next download
+ - null: External storage, but cache status not checked (use include_cache_status=true)
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ path : typing.Optional[str]
+ Filter files by path prefix
+
+ include_cache_status : typing.Optional[bool]
+ Check and return cache status for each file. When false, storage files return null for cache_status.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ListFilesetFilesResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets/{encode_path_param(name)}/files",
+ method="GET",
+ params={
+ "path": path,
+ "include_cache_status": include_cache_status,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ListFilesetFilesResponse,
+ parse_obj_as(
+ type_=ListFilesetFilesResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
diff --git a/sdks/python/files/types/__init__.py b/sdks/python/files/types/__init__.py
new file mode 100644
index 0000000000..93bfae4215
--- /dev/null
+++ b/sdks/python/files/types/__init__.py
@@ -0,0 +1,52 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
+import typing
+from importlib import import_module
+
+if typing.TYPE_CHECKING:
+ from .create_fileset_request_storage import (
+ CreateFilesetRequestStorage,
+ CreateFilesetRequestStorage_Huggingface,
+ CreateFilesetRequestStorage_Local,
+ CreateFilesetRequestStorage_Ngc,
+ CreateFilesetRequestStorage_S3,
+ )
+_dynamic_imports: typing.Dict[str, str] = {
+ "CreateFilesetRequestStorage": ".create_fileset_request_storage",
+ "CreateFilesetRequestStorage_Huggingface": ".create_fileset_request_storage",
+ "CreateFilesetRequestStorage_Local": ".create_fileset_request_storage",
+ "CreateFilesetRequestStorage_Ngc": ".create_fileset_request_storage",
+ "CreateFilesetRequestStorage_S3": ".create_fileset_request_storage",
+}
+
+
+def __getattr__(attr_name: str) -> typing.Any:
+ module_name = _dynamic_imports.get(attr_name)
+ if module_name is None:
+ raise AttributeError(f"No {attr_name} found in _dynamic_imports for module name -> {__name__}")
+ try:
+ module = import_module(module_name, __package__)
+ if module_name == f".{attr_name}":
+ return module
+ else:
+ return getattr(module, attr_name)
+ except ImportError as e:
+ raise ImportError(f"Failed to import {attr_name} from {module_name}: {e}") from e
+ except AttributeError as e:
+ raise AttributeError(f"Failed to get {attr_name} from {module_name}: {e}") from e
+
+
+def __dir__():
+ lazy_attrs = list(_dynamic_imports.keys())
+ return sorted(lazy_attrs)
+
+
+__all__ = [
+ "CreateFilesetRequestStorage",
+ "CreateFilesetRequestStorage_Huggingface",
+ "CreateFilesetRequestStorage_Local",
+ "CreateFilesetRequestStorage_Ngc",
+ "CreateFilesetRequestStorage_S3",
+]
diff --git a/sdks/python/files/types/create_fileset_request_storage.py b/sdks/python/files/types/create_fileset_request_storage.py
new file mode 100644
index 0000000000..5a2720bb29
--- /dev/null
+++ b/sdks/python/files/types/create_fileset_request_storage.py
@@ -0,0 +1,120 @@
+# This file was auto-generated by Fern from our API Definition.
+
+from __future__ import annotations
+
+import typing
+
+import pydantic
+import typing_extensions
+from ...core.pydantic_utilities import IS_PYDANTIC_V2, UniversalBaseModel
+from ...types.huggingface_storage_config_repo_type import HuggingfaceStorageConfigRepoType
+from ...types.ngc_storage_config_target_type import NgcStorageConfigTargetType
+from ...types.s3storage_config_signature_version import S3StorageConfigSignatureVersion
+from ...types.secret_ref import SecretRef
+
+
+class CreateFilesetRequestStorage_Local(UniversalBaseModel):
+ """
+ The storage configuration for the fileset. If not provided, uses default storage.
+ """
+
+ type: typing.Literal["local"] = "local"
+ read_chunk_size: typing.Optional[int] = None
+ path: str
+ write_buffer_size: typing.Optional[int] = None
+
+ if IS_PYDANTIC_V2:
+ model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True) # type: ignore # Pydantic v2
+ else:
+
+ class Config:
+ frozen = True
+ smart_union = True
+ extra = pydantic.Extra.allow
+
+
+class CreateFilesetRequestStorage_Ngc(UniversalBaseModel):
+ """
+ The storage configuration for the fileset. If not provided, uses default storage.
+ """
+
+ type: typing.Literal["ngc"] = "ngc"
+ read_chunk_size: typing.Optional[int] = None
+ org: str
+ team: str
+ target: str
+ target_type: typing.Optional[NgcStorageConfigTargetType] = None
+ version: typing.Optional[str] = None
+ original_version: typing.Optional[str] = None
+ api_key_secret: SecretRef
+ host: typing.Optional[str] = None
+
+ if IS_PYDANTIC_V2:
+ model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True) # type: ignore # Pydantic v2
+ else:
+
+ class Config:
+ frozen = True
+ smart_union = True
+ extra = pydantic.Extra.allow
+
+
+class CreateFilesetRequestStorage_Huggingface(UniversalBaseModel):
+ """
+ The storage configuration for the fileset. If not provided, uses default storage.
+ """
+
+ type: typing.Literal["huggingface"] = "huggingface"
+ read_chunk_size: typing.Optional[int] = None
+ repo_id: str
+ repo_type: typing.Optional[HuggingfaceStorageConfigRepoType] = None
+ revision: typing.Optional[str] = None
+ original_revision: typing.Optional[str] = None
+ token_secret: typing.Optional[SecretRef] = None
+ endpoint: typing.Optional[str] = None
+
+ if IS_PYDANTIC_V2:
+ model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True) # type: ignore # Pydantic v2
+ else:
+
+ class Config:
+ frozen = True
+ smart_union = True
+ extra = pydantic.Extra.allow
+
+
+class CreateFilesetRequestStorage_S3(UniversalBaseModel):
+ """
+ The storage configuration for the fileset. If not provided, uses default storage.
+ """
+
+ type: typing.Literal["s3"] = "s3"
+ read_chunk_size: typing.Optional[int] = None
+ bucket: str
+ prefix: typing.Optional[str] = None
+ region: typing.Optional[str] = None
+ endpoint_url: typing.Optional[str] = None
+ use_sdk_auth: typing.Optional[bool] = None
+ access_key_id_secret: typing.Optional[SecretRef] = None
+ secret_access_key_secret: typing.Optional[SecretRef] = None
+ signature_version: typing.Optional[S3StorageConfigSignatureVersion] = None
+
+ if IS_PYDANTIC_V2:
+ model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True) # type: ignore # Pydantic v2
+ else:
+
+ class Config:
+ frozen = True
+ smart_union = True
+ extra = pydantic.Extra.allow
+
+
+CreateFilesetRequestStorage = typing_extensions.Annotated[
+ typing.Union[
+ CreateFilesetRequestStorage_Local,
+ CreateFilesetRequestStorage_Ngc,
+ CreateFilesetRequestStorage_Huggingface,
+ CreateFilesetRequestStorage_S3,
+ ],
+ pydantic.Field(discriminator="type"),
+]
diff --git a/sdks/python/guardrails/__init__.py b/sdks/python/guardrails/__init__.py
new file mode 100644
index 0000000000..54fa462621
--- /dev/null
+++ b/sdks/python/guardrails/__init__.py
@@ -0,0 +1,64 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
+import typing
+from importlib import import_module
+
+if typing.TYPE_CHECKING:
+ from .types import (
+ GuardrailCheckRequestFunctionCall,
+ GuardrailCheckRequestMessagesItem,
+ GuardrailCheckRequestMessagesItem_Assistant,
+ GuardrailCheckRequestMessagesItem_Function,
+ GuardrailCheckRequestMessagesItem_System,
+ GuardrailCheckRequestMessagesItem_Tool,
+ GuardrailCheckRequestMessagesItem_User,
+ GuardrailCheckRequestStop,
+ GuardrailCheckRequestToolChoice,
+ )
+_dynamic_imports: typing.Dict[str, str] = {
+ "GuardrailCheckRequestFunctionCall": ".types",
+ "GuardrailCheckRequestMessagesItem": ".types",
+ "GuardrailCheckRequestMessagesItem_Assistant": ".types",
+ "GuardrailCheckRequestMessagesItem_Function": ".types",
+ "GuardrailCheckRequestMessagesItem_System": ".types",
+ "GuardrailCheckRequestMessagesItem_Tool": ".types",
+ "GuardrailCheckRequestMessagesItem_User": ".types",
+ "GuardrailCheckRequestStop": ".types",
+ "GuardrailCheckRequestToolChoice": ".types",
+}
+
+
+def __getattr__(attr_name: str) -> typing.Any:
+ module_name = _dynamic_imports.get(attr_name)
+ if module_name is None:
+ raise AttributeError(f"No {attr_name} found in _dynamic_imports for module name -> {__name__}")
+ try:
+ module = import_module(module_name, __package__)
+ if module_name == f".{attr_name}":
+ return module
+ else:
+ return getattr(module, attr_name)
+ except ImportError as e:
+ raise ImportError(f"Failed to import {attr_name} from {module_name}: {e}") from e
+ except AttributeError as e:
+ raise AttributeError(f"Failed to get {attr_name} from {module_name}: {e}") from e
+
+
+def __dir__():
+ lazy_attrs = list(_dynamic_imports.keys())
+ return sorted(lazy_attrs)
+
+
+__all__ = [
+ "GuardrailCheckRequestFunctionCall",
+ "GuardrailCheckRequestMessagesItem",
+ "GuardrailCheckRequestMessagesItem_Assistant",
+ "GuardrailCheckRequestMessagesItem_Function",
+ "GuardrailCheckRequestMessagesItem_System",
+ "GuardrailCheckRequestMessagesItem_Tool",
+ "GuardrailCheckRequestMessagesItem_User",
+ "GuardrailCheckRequestStop",
+ "GuardrailCheckRequestToolChoice",
+]
diff --git a/sdks/python/guardrails/client.py b/sdks/python/guardrails/client.py
new file mode 100644
index 0000000000..2490840caf
--- /dev/null
+++ b/sdks/python/guardrails/client.py
@@ -0,0 +1,891 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.request_options import RequestOptions
+from ..types.delete_response import DeleteResponse
+from ..types.generic_sort_field import GenericSortField
+from ..types.guardrail_check_response import GuardrailCheckResponse
+from ..types.guardrail_config import GuardrailConfig
+from ..types.guardrail_config_filter import GuardrailConfigFilter
+from ..types.guardrail_configs_page import GuardrailConfigsPage
+from ..types.guardrails_data_input import GuardrailsDataInput
+from .raw_client import AsyncRawGuardrailsClient, RawGuardrailsClient
+from .types.guardrail_check_request_function_call import GuardrailCheckRequestFunctionCall
+from .types.guardrail_check_request_messages_item import GuardrailCheckRequestMessagesItem
+from .types.guardrail_check_request_stop import GuardrailCheckRequestStop
+from .types.guardrail_check_request_tool_choice import GuardrailCheckRequestToolChoice
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class GuardrailsClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._raw_client = RawGuardrailsClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> RawGuardrailsClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ RawGuardrailsClient
+ """
+ return self._raw_client
+
+ def check(
+ self,
+ workspace: str,
+ *,
+ model: str,
+ messages: typing.Sequence[GuardrailCheckRequestMessagesItem],
+ response_format: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ max_tokens: typing.Optional[int] = OMIT,
+ n: typing.Optional[int] = OMIT,
+ stream: typing.Optional[bool] = OMIT,
+ temperature: typing.Optional[float] = OMIT,
+ top_p: typing.Optional[float] = OMIT,
+ stop: typing.Optional[GuardrailCheckRequestStop] = OMIT,
+ frequency_penalty: typing.Optional[float] = OMIT,
+ presence_penalty: typing.Optional[float] = OMIT,
+ function_call: typing.Optional[GuardrailCheckRequestFunctionCall] = OMIT,
+ seed: typing.Optional[int] = OMIT,
+ logit_bias: typing.Optional[typing.Dict[str, float]] = OMIT,
+ top_logprobs: typing.Optional[int] = OMIT,
+ logprobs: typing.Optional[bool] = OMIT,
+ tool_choice: typing.Optional[GuardrailCheckRequestToolChoice] = OMIT,
+ user: typing.Optional[str] = OMIT,
+ tools: typing.Optional[typing.Sequence[typing.Dict[str, typing.Any]]] = OMIT,
+ ignore_eos: typing.Optional[bool] = OMIT,
+ reasoning_effort: typing.Optional[str] = OMIT,
+ max_completion_tokens: typing.Optional[int] = OMIT,
+ stream_options: typing.Optional[typing.Dict[str, bool]] = OMIT,
+ vision: typing.Optional[bool] = OMIT,
+ guardrails: typing.Optional[GuardrailsDataInput] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> GuardrailCheckResponse:
+ """
+ Chat completion for the provided conversation.
+
+ Parameters
+ ----------
+ workspace : str
+
+ model : str
+ The model to use for completion. Must be one of the available models.
+
+ messages : typing.Sequence[GuardrailCheckRequestMessagesItem]
+ A list of messages comprising the conversation so far
+
+ response_format : typing.Optional[typing.Dict[str, typing.Any]]
+ Format of the response. Use {'type': 'json_object'} for JSON mode or {'type': 'json_schema', 'json_schema': {...}} for structured outputs.
+
+ max_tokens : typing.Optional[int]
+ The maximum number of tokens that can be generated in the chat completion.
+
+ n : typing.Optional[int]
+ How many chat completion choices to generate for each input message.
+
+ stream : typing.Optional[bool]
+ If set, partial message deltas will be sent, like in ChatGPT.
+
+ temperature : typing.Optional[float]
+ What sampling temperature to use, between 0 and 2.
+
+ top_p : typing.Optional[float]
+ An alternative to sampling with temperature, called nucleus sampling.
+
+ stop : typing.Optional[GuardrailCheckRequestStop]
+ Up to 4 sequences where the API will stop generating further tokens.
+
+ frequency_penalty : typing.Optional[float]
+ Positive values penalize new tokens based on their existing frequency in the text.
+
+ presence_penalty : typing.Optional[float]
+ Positive values penalize new tokens based on whether they appear in the text so far.
+
+ function_call : typing.Optional[GuardrailCheckRequestFunctionCall]
+ Deprecated in favor of tool_choice. 'none' means the model will not call a function and instead generates a message. 'auto' means the model can pick between generating a message or calling a function. Specifying a particular function via {'name': 'my_function'} forces the model to call that function.
+
+ seed : typing.Optional[int]
+ If specified, attempts to sample deterministically.
+
+ logit_bias : typing.Optional[typing.Dict[str, float]]
+ Modify the likelihood of specified tokens appearing in the completion. Maps token IDs (as strings) to bias values from -100 to 100.
+
+ top_logprobs : typing.Optional[int]
+ The number of most likely tokens to return at each token position.
+
+ logprobs : typing.Optional[bool]
+ Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message
+
+ tool_choice : typing.Optional[GuardrailCheckRequestToolChoice]
+ Controls which (if any) tool is called by the model. 'none' means no tool is called, 'auto' lets the model decide, 'required' forces a tool call.
+
+ user : typing.Optional[str]
+ A unique identifier representing your end-user, used by some providers for abuse monitoring.
+
+ tools : typing.Optional[typing.Sequence[typing.Dict[str, typing.Any]]]
+ A list of tools the model may call. Each tool is an object with a 'type' field and a 'function' definition.
+
+ ignore_eos : typing.Optional[bool]
+ Ignore the eos when running
+
+ reasoning_effort : typing.Optional[str]
+ Constrains effort on reasoning for reasoning models. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.
+
+ max_completion_tokens : typing.Optional[int]
+ An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens. Preferred over max_tokens for reasoning models.
+
+ stream_options : typing.Optional[typing.Dict[str, bool]]
+ Options for streaming response. Only set this when stream=True. Supports include_usage to receive token usage in the final stream chunk.
+
+ vision : typing.Optional[bool]
+ Whether this is a vision-capable request with image inputs.
+
+ guardrails : typing.Optional[GuardrailsDataInput]
+ Guardrails specific options for the request.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ GuardrailCheckResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+ from nvidia.guardrails import GuardrailCheckRequestMessagesItem_System
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.guardrails.check(
+ workspace="workspace",
+ model="model",
+ messages=[
+ GuardrailCheckRequestMessagesItem_System(
+ content="content",
+ )
+ ],
+ )
+ """
+ _response = self._raw_client.check(
+ workspace,
+ model=model,
+ messages=messages,
+ response_format=response_format,
+ max_tokens=max_tokens,
+ n=n,
+ stream=stream,
+ temperature=temperature,
+ top_p=top_p,
+ stop=stop,
+ frequency_penalty=frequency_penalty,
+ presence_penalty=presence_penalty,
+ function_call=function_call,
+ seed=seed,
+ logit_bias=logit_bias,
+ top_logprobs=top_logprobs,
+ logprobs=logprobs,
+ tool_choice=tool_choice,
+ user=user,
+ tools=tools,
+ ignore_eos=ignore_eos,
+ reasoning_effort=reasoning_effort,
+ max_completion_tokens=max_completion_tokens,
+ stream_options=stream_options,
+ vision=vision,
+ guardrails=guardrails,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def list_guardrail_configs(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[GenericSortField] = None,
+ filter: typing.Optional[GuardrailConfigFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> GuardrailConfigsPage:
+ """
+ List available guardrail configs.
+
+ Lists guardrail configs for a specific workspace.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[GenericSortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[GuardrailConfigFilter]
+ Filter guardrail configs by name, description, project, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ GuardrailConfigsPage
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.guardrails.list_guardrail_configs(
+ workspace="workspace",
+ )
+ """
+ _response = self._raw_client.list_guardrail_configs(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ def create_config(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ description: typing.Optional[str] = OMIT,
+ data: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> GuardrailConfig:
+ """
+ Create a new guardrail config.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ The name of the guardrail config
+
+ description : typing.Optional[str]
+ Description of the guardrail config
+
+ data : typing.Optional[typing.Dict[str, typing.Any]]
+ Guardrail configuration data
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ GuardrailConfig
+ Config created successfully.
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.guardrails.create_config(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.create_config(
+ workspace, name=name, description=description, data=data, request_options=request_options
+ )
+ return _response.data
+
+ def get_guardrail_config(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> GuardrailConfig:
+ """
+ Get info about a guardrail configuration.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ GuardrailConfig
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.guardrails.get_guardrail_config(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.get_guardrail_config(workspace, name, request_options=request_options)
+ return _response.data
+
+ def delete_config(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> DeleteResponse:
+ """
+ Delete a guardrail config.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ DeleteResponse
+ Successful model deletion.
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.guardrails.delete_config(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.delete_config(workspace, name, request_options=request_options)
+ return _response.data
+
+ def update_config(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ data: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> GuardrailConfig:
+ """
+ Update model metadata. If the request body has an empty field,
+ keep the old value.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ description : typing.Optional[str]
+ Description of the guardrail config
+
+ data : typing.Optional[typing.Dict[str, typing.Any]]
+ Guardrail configuration data
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ GuardrailConfig
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.guardrails.update_config(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.update_config(
+ workspace, name, description=description, data=data, request_options=request_options
+ )
+ return _response.data
+
+
+class AsyncGuardrailsClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._raw_client = AsyncRawGuardrailsClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> AsyncRawGuardrailsClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ AsyncRawGuardrailsClient
+ """
+ return self._raw_client
+
+ async def check(
+ self,
+ workspace: str,
+ *,
+ model: str,
+ messages: typing.Sequence[GuardrailCheckRequestMessagesItem],
+ response_format: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ max_tokens: typing.Optional[int] = OMIT,
+ n: typing.Optional[int] = OMIT,
+ stream: typing.Optional[bool] = OMIT,
+ temperature: typing.Optional[float] = OMIT,
+ top_p: typing.Optional[float] = OMIT,
+ stop: typing.Optional[GuardrailCheckRequestStop] = OMIT,
+ frequency_penalty: typing.Optional[float] = OMIT,
+ presence_penalty: typing.Optional[float] = OMIT,
+ function_call: typing.Optional[GuardrailCheckRequestFunctionCall] = OMIT,
+ seed: typing.Optional[int] = OMIT,
+ logit_bias: typing.Optional[typing.Dict[str, float]] = OMIT,
+ top_logprobs: typing.Optional[int] = OMIT,
+ logprobs: typing.Optional[bool] = OMIT,
+ tool_choice: typing.Optional[GuardrailCheckRequestToolChoice] = OMIT,
+ user: typing.Optional[str] = OMIT,
+ tools: typing.Optional[typing.Sequence[typing.Dict[str, typing.Any]]] = OMIT,
+ ignore_eos: typing.Optional[bool] = OMIT,
+ reasoning_effort: typing.Optional[str] = OMIT,
+ max_completion_tokens: typing.Optional[int] = OMIT,
+ stream_options: typing.Optional[typing.Dict[str, bool]] = OMIT,
+ vision: typing.Optional[bool] = OMIT,
+ guardrails: typing.Optional[GuardrailsDataInput] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> GuardrailCheckResponse:
+ """
+ Chat completion for the provided conversation.
+
+ Parameters
+ ----------
+ workspace : str
+
+ model : str
+ The model to use for completion. Must be one of the available models.
+
+ messages : typing.Sequence[GuardrailCheckRequestMessagesItem]
+ A list of messages comprising the conversation so far
+
+ response_format : typing.Optional[typing.Dict[str, typing.Any]]
+ Format of the response. Use {'type': 'json_object'} for JSON mode or {'type': 'json_schema', 'json_schema': {...}} for structured outputs.
+
+ max_tokens : typing.Optional[int]
+ The maximum number of tokens that can be generated in the chat completion.
+
+ n : typing.Optional[int]
+ How many chat completion choices to generate for each input message.
+
+ stream : typing.Optional[bool]
+ If set, partial message deltas will be sent, like in ChatGPT.
+
+ temperature : typing.Optional[float]
+ What sampling temperature to use, between 0 and 2.
+
+ top_p : typing.Optional[float]
+ An alternative to sampling with temperature, called nucleus sampling.
+
+ stop : typing.Optional[GuardrailCheckRequestStop]
+ Up to 4 sequences where the API will stop generating further tokens.
+
+ frequency_penalty : typing.Optional[float]
+ Positive values penalize new tokens based on their existing frequency in the text.
+
+ presence_penalty : typing.Optional[float]
+ Positive values penalize new tokens based on whether they appear in the text so far.
+
+ function_call : typing.Optional[GuardrailCheckRequestFunctionCall]
+ Deprecated in favor of tool_choice. 'none' means the model will not call a function and instead generates a message. 'auto' means the model can pick between generating a message or calling a function. Specifying a particular function via {'name': 'my_function'} forces the model to call that function.
+
+ seed : typing.Optional[int]
+ If specified, attempts to sample deterministically.
+
+ logit_bias : typing.Optional[typing.Dict[str, float]]
+ Modify the likelihood of specified tokens appearing in the completion. Maps token IDs (as strings) to bias values from -100 to 100.
+
+ top_logprobs : typing.Optional[int]
+ The number of most likely tokens to return at each token position.
+
+ logprobs : typing.Optional[bool]
+ Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message
+
+ tool_choice : typing.Optional[GuardrailCheckRequestToolChoice]
+ Controls which (if any) tool is called by the model. 'none' means no tool is called, 'auto' lets the model decide, 'required' forces a tool call.
+
+ user : typing.Optional[str]
+ A unique identifier representing your end-user, used by some providers for abuse monitoring.
+
+ tools : typing.Optional[typing.Sequence[typing.Dict[str, typing.Any]]]
+ A list of tools the model may call. Each tool is an object with a 'type' field and a 'function' definition.
+
+ ignore_eos : typing.Optional[bool]
+ Ignore the eos when running
+
+ reasoning_effort : typing.Optional[str]
+ Constrains effort on reasoning for reasoning models. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.
+
+ max_completion_tokens : typing.Optional[int]
+ An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens. Preferred over max_tokens for reasoning models.
+
+ stream_options : typing.Optional[typing.Dict[str, bool]]
+ Options for streaming response. Only set this when stream=True. Supports include_usage to receive token usage in the final stream chunk.
+
+ vision : typing.Optional[bool]
+ Whether this is a vision-capable request with image inputs.
+
+ guardrails : typing.Optional[GuardrailsDataInput]
+ Guardrails specific options for the request.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ GuardrailCheckResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+ from nvidia.guardrails import GuardrailCheckRequestMessagesItem_System
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.guardrails.check(
+ workspace="workspace",
+ model="model",
+ messages=[
+ GuardrailCheckRequestMessagesItem_System(
+ content="content",
+ )
+ ],
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.check(
+ workspace,
+ model=model,
+ messages=messages,
+ response_format=response_format,
+ max_tokens=max_tokens,
+ n=n,
+ stream=stream,
+ temperature=temperature,
+ top_p=top_p,
+ stop=stop,
+ frequency_penalty=frequency_penalty,
+ presence_penalty=presence_penalty,
+ function_call=function_call,
+ seed=seed,
+ logit_bias=logit_bias,
+ top_logprobs=top_logprobs,
+ logprobs=logprobs,
+ tool_choice=tool_choice,
+ user=user,
+ tools=tools,
+ ignore_eos=ignore_eos,
+ reasoning_effort=reasoning_effort,
+ max_completion_tokens=max_completion_tokens,
+ stream_options=stream_options,
+ vision=vision,
+ guardrails=guardrails,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def list_guardrail_configs(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[GenericSortField] = None,
+ filter: typing.Optional[GuardrailConfigFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> GuardrailConfigsPage:
+ """
+ List available guardrail configs.
+
+ Lists guardrail configs for a specific workspace.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[GenericSortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[GuardrailConfigFilter]
+ Filter guardrail configs by name, description, project, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ GuardrailConfigsPage
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.guardrails.list_guardrail_configs(
+ workspace="workspace",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_guardrail_configs(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ async def create_config(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ description: typing.Optional[str] = OMIT,
+ data: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> GuardrailConfig:
+ """
+ Create a new guardrail config.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ The name of the guardrail config
+
+ description : typing.Optional[str]
+ Description of the guardrail config
+
+ data : typing.Optional[typing.Dict[str, typing.Any]]
+ Guardrail configuration data
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ GuardrailConfig
+ Config created successfully.
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.guardrails.create_config(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.create_config(
+ workspace, name=name, description=description, data=data, request_options=request_options
+ )
+ return _response.data
+
+ async def get_guardrail_config(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> GuardrailConfig:
+ """
+ Get info about a guardrail configuration.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ GuardrailConfig
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.guardrails.get_guardrail_config(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_guardrail_config(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def delete_config(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> DeleteResponse:
+ """
+ Delete a guardrail config.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ DeleteResponse
+ Successful model deletion.
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.guardrails.delete_config(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.delete_config(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def update_config(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ data: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> GuardrailConfig:
+ """
+ Update model metadata. If the request body has an empty field,
+ keep the old value.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ description : typing.Optional[str]
+ Description of the guardrail config
+
+ data : typing.Optional[typing.Dict[str, typing.Any]]
+ Guardrail configuration data
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ GuardrailConfig
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.guardrails.update_config(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.update_config(
+ workspace, name, description=description, data=data, request_options=request_options
+ )
+ return _response.data
diff --git a/sdks/python/guardrails/raw_client.py b/sdks/python/guardrails/raw_client.py
new file mode 100644
index 0000000000..ee6890888c
--- /dev/null
+++ b/sdks/python/guardrails/raw_client.py
@@ -0,0 +1,1259 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+from json.decoder import JSONDecodeError
+
+from ..core.api_error import ApiError
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.http_response import AsyncHttpResponse, HttpResponse
+from ..core.jsonable_encoder import encode_path_param
+from ..core.parse_error import ParsingError
+from ..core.pydantic_utilities import parse_obj_as
+from ..core.request_options import RequestOptions
+from ..core.serialization import convert_and_respect_annotation_metadata
+from ..errors.bad_request_error import BadRequestError
+from ..errors.conflict_error import ConflictError
+from ..errors.internal_server_error import InternalServerError
+from ..errors.not_found_error import NotFoundError
+from ..errors.unprocessable_entity_error import UnprocessableEntityError
+from ..types.delete_response import DeleteResponse
+from ..types.generic_sort_field import GenericSortField
+from ..types.guardrail_check_response import GuardrailCheckResponse
+from ..types.guardrail_config import GuardrailConfig
+from ..types.guardrail_config_filter import GuardrailConfigFilter
+from ..types.guardrail_configs_page import GuardrailConfigsPage
+from ..types.guardrails_data_input import GuardrailsDataInput
+from .types.guardrail_check_request_function_call import GuardrailCheckRequestFunctionCall
+from .types.guardrail_check_request_messages_item import GuardrailCheckRequestMessagesItem
+from .types.guardrail_check_request_stop import GuardrailCheckRequestStop
+from .types.guardrail_check_request_tool_choice import GuardrailCheckRequestToolChoice
+from pydantic import ValidationError
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class RawGuardrailsClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ def check(
+ self,
+ workspace: str,
+ *,
+ model: str,
+ messages: typing.Sequence[GuardrailCheckRequestMessagesItem],
+ response_format: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ max_tokens: typing.Optional[int] = OMIT,
+ n: typing.Optional[int] = OMIT,
+ stream: typing.Optional[bool] = OMIT,
+ temperature: typing.Optional[float] = OMIT,
+ top_p: typing.Optional[float] = OMIT,
+ stop: typing.Optional[GuardrailCheckRequestStop] = OMIT,
+ frequency_penalty: typing.Optional[float] = OMIT,
+ presence_penalty: typing.Optional[float] = OMIT,
+ function_call: typing.Optional[GuardrailCheckRequestFunctionCall] = OMIT,
+ seed: typing.Optional[int] = OMIT,
+ logit_bias: typing.Optional[typing.Dict[str, float]] = OMIT,
+ top_logprobs: typing.Optional[int] = OMIT,
+ logprobs: typing.Optional[bool] = OMIT,
+ tool_choice: typing.Optional[GuardrailCheckRequestToolChoice] = OMIT,
+ user: typing.Optional[str] = OMIT,
+ tools: typing.Optional[typing.Sequence[typing.Dict[str, typing.Any]]] = OMIT,
+ ignore_eos: typing.Optional[bool] = OMIT,
+ reasoning_effort: typing.Optional[str] = OMIT,
+ max_completion_tokens: typing.Optional[int] = OMIT,
+ stream_options: typing.Optional[typing.Dict[str, bool]] = OMIT,
+ vision: typing.Optional[bool] = OMIT,
+ guardrails: typing.Optional[GuardrailsDataInput] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[GuardrailCheckResponse]:
+ """
+ Chat completion for the provided conversation.
+
+ Parameters
+ ----------
+ workspace : str
+
+ model : str
+ The model to use for completion. Must be one of the available models.
+
+ messages : typing.Sequence[GuardrailCheckRequestMessagesItem]
+ A list of messages comprising the conversation so far
+
+ response_format : typing.Optional[typing.Dict[str, typing.Any]]
+ Format of the response. Use {'type': 'json_object'} for JSON mode or {'type': 'json_schema', 'json_schema': {...}} for structured outputs.
+
+ max_tokens : typing.Optional[int]
+ The maximum number of tokens that can be generated in the chat completion.
+
+ n : typing.Optional[int]
+ How many chat completion choices to generate for each input message.
+
+ stream : typing.Optional[bool]
+ If set, partial message deltas will be sent, like in ChatGPT.
+
+ temperature : typing.Optional[float]
+ What sampling temperature to use, between 0 and 2.
+
+ top_p : typing.Optional[float]
+ An alternative to sampling with temperature, called nucleus sampling.
+
+ stop : typing.Optional[GuardrailCheckRequestStop]
+ Up to 4 sequences where the API will stop generating further tokens.
+
+ frequency_penalty : typing.Optional[float]
+ Positive values penalize new tokens based on their existing frequency in the text.
+
+ presence_penalty : typing.Optional[float]
+ Positive values penalize new tokens based on whether they appear in the text so far.
+
+ function_call : typing.Optional[GuardrailCheckRequestFunctionCall]
+ Deprecated in favor of tool_choice. 'none' means the model will not call a function and instead generates a message. 'auto' means the model can pick between generating a message or calling a function. Specifying a particular function via {'name': 'my_function'} forces the model to call that function.
+
+ seed : typing.Optional[int]
+ If specified, attempts to sample deterministically.
+
+ logit_bias : typing.Optional[typing.Dict[str, float]]
+ Modify the likelihood of specified tokens appearing in the completion. Maps token IDs (as strings) to bias values from -100 to 100.
+
+ top_logprobs : typing.Optional[int]
+ The number of most likely tokens to return at each token position.
+
+ logprobs : typing.Optional[bool]
+ Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message
+
+ tool_choice : typing.Optional[GuardrailCheckRequestToolChoice]
+ Controls which (if any) tool is called by the model. 'none' means no tool is called, 'auto' lets the model decide, 'required' forces a tool call.
+
+ user : typing.Optional[str]
+ A unique identifier representing your end-user, used by some providers for abuse monitoring.
+
+ tools : typing.Optional[typing.Sequence[typing.Dict[str, typing.Any]]]
+ A list of tools the model may call. Each tool is an object with a 'type' field and a 'function' definition.
+
+ ignore_eos : typing.Optional[bool]
+ Ignore the eos when running
+
+ reasoning_effort : typing.Optional[str]
+ Constrains effort on reasoning for reasoning models. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.
+
+ max_completion_tokens : typing.Optional[int]
+ An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens. Preferred over max_tokens for reasoning models.
+
+ stream_options : typing.Optional[typing.Dict[str, bool]]
+ Options for streaming response. Only set this when stream=True. Supports include_usage to receive token usage in the final stream chunk.
+
+ vision : typing.Optional[bool]
+ Whether this is a vision-capable request with image inputs.
+
+ guardrails : typing.Optional[GuardrailsDataInput]
+ Guardrails specific options for the request.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[GuardrailCheckResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/guardrails/v2/workspaces/{encode_path_param(workspace)}/checks",
+ method="POST",
+ json={
+ "model": model,
+ "response_format": response_format,
+ "max_tokens": max_tokens,
+ "n": n,
+ "stream": stream,
+ "temperature": temperature,
+ "top_p": top_p,
+ "stop": convert_and_respect_annotation_metadata(
+ object_=stop, annotation=GuardrailCheckRequestStop, direction="write"
+ ),
+ "frequency_penalty": frequency_penalty,
+ "presence_penalty": presence_penalty,
+ "function_call": convert_and_respect_annotation_metadata(
+ object_=function_call, annotation=GuardrailCheckRequestFunctionCall, direction="write"
+ ),
+ "seed": seed,
+ "logit_bias": logit_bias,
+ "top_logprobs": top_logprobs,
+ "logprobs": logprobs,
+ "tool_choice": convert_and_respect_annotation_metadata(
+ object_=tool_choice, annotation=GuardrailCheckRequestToolChoice, direction="write"
+ ),
+ "user": user,
+ "tools": tools,
+ "ignore_eos": ignore_eos,
+ "reasoning_effort": reasoning_effort,
+ "max_completion_tokens": max_completion_tokens,
+ "stream_options": stream_options,
+ "messages": convert_and_respect_annotation_metadata(
+ object_=messages, annotation=typing.Sequence[GuardrailCheckRequestMessagesItem], direction="write"
+ ),
+ "vision": vision,
+ "guardrails": convert_and_respect_annotation_metadata(
+ object_=guardrails, annotation=GuardrailsDataInput, direction="write"
+ ),
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ GuardrailCheckResponse,
+ parse_obj_as(
+ type_=GuardrailCheckResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 400:
+ raise BadRequestError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 500:
+ raise InternalServerError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def list_guardrail_configs(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[GenericSortField] = None,
+ filter: typing.Optional[GuardrailConfigFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[GuardrailConfigsPage]:
+ """
+ List available guardrail configs.
+
+ Lists guardrail configs for a specific workspace.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[GenericSortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[GuardrailConfigFilter]
+ Filter guardrail configs by name, description, project, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[GuardrailConfigsPage]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/guardrails/v2/workspaces/{encode_path_param(workspace)}/configs",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=GuardrailConfigFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ GuardrailConfigsPage,
+ parse_obj_as(
+ type_=GuardrailConfigsPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def create_config(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ description: typing.Optional[str] = OMIT,
+ data: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[GuardrailConfig]:
+ """
+ Create a new guardrail config.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ The name of the guardrail config
+
+ description : typing.Optional[str]
+ Description of the guardrail config
+
+ data : typing.Optional[typing.Dict[str, typing.Any]]
+ Guardrail configuration data
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[GuardrailConfig]
+ Config created successfully.
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/guardrails/v2/workspaces/{encode_path_param(workspace)}/configs",
+ method="POST",
+ json={
+ "name": name,
+ "description": description,
+ "data": data,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ GuardrailConfig,
+ parse_obj_as(
+ type_=GuardrailConfig, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 409:
+ raise ConflictError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_guardrail_config(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[GuardrailConfig]:
+ """
+ Get info about a guardrail configuration.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[GuardrailConfig]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/guardrails/v2/workspaces/{encode_path_param(workspace)}/configs/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ GuardrailConfig,
+ parse_obj_as(
+ type_=GuardrailConfig, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def delete_config(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[DeleteResponse]:
+ """
+ Delete a guardrail config.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[DeleteResponse]
+ Successful model deletion.
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/guardrails/v2/workspaces/{encode_path_param(workspace)}/configs/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ DeleteResponse,
+ parse_obj_as(
+ type_=DeleteResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def update_config(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ data: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[GuardrailConfig]:
+ """
+ Update model metadata. If the request body has an empty field,
+ keep the old value.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ description : typing.Optional[str]
+ Description of the guardrail config
+
+ data : typing.Optional[typing.Dict[str, typing.Any]]
+ Guardrail configuration data
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[GuardrailConfig]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/guardrails/v2/workspaces/{encode_path_param(workspace)}/configs/{encode_path_param(name)}",
+ method="PATCH",
+ json={
+ "description": description,
+ "data": data,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ GuardrailConfig,
+ parse_obj_as(
+ type_=GuardrailConfig, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+
+class AsyncRawGuardrailsClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ async def check(
+ self,
+ workspace: str,
+ *,
+ model: str,
+ messages: typing.Sequence[GuardrailCheckRequestMessagesItem],
+ response_format: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ max_tokens: typing.Optional[int] = OMIT,
+ n: typing.Optional[int] = OMIT,
+ stream: typing.Optional[bool] = OMIT,
+ temperature: typing.Optional[float] = OMIT,
+ top_p: typing.Optional[float] = OMIT,
+ stop: typing.Optional[GuardrailCheckRequestStop] = OMIT,
+ frequency_penalty: typing.Optional[float] = OMIT,
+ presence_penalty: typing.Optional[float] = OMIT,
+ function_call: typing.Optional[GuardrailCheckRequestFunctionCall] = OMIT,
+ seed: typing.Optional[int] = OMIT,
+ logit_bias: typing.Optional[typing.Dict[str, float]] = OMIT,
+ top_logprobs: typing.Optional[int] = OMIT,
+ logprobs: typing.Optional[bool] = OMIT,
+ tool_choice: typing.Optional[GuardrailCheckRequestToolChoice] = OMIT,
+ user: typing.Optional[str] = OMIT,
+ tools: typing.Optional[typing.Sequence[typing.Dict[str, typing.Any]]] = OMIT,
+ ignore_eos: typing.Optional[bool] = OMIT,
+ reasoning_effort: typing.Optional[str] = OMIT,
+ max_completion_tokens: typing.Optional[int] = OMIT,
+ stream_options: typing.Optional[typing.Dict[str, bool]] = OMIT,
+ vision: typing.Optional[bool] = OMIT,
+ guardrails: typing.Optional[GuardrailsDataInput] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[GuardrailCheckResponse]:
+ """
+ Chat completion for the provided conversation.
+
+ Parameters
+ ----------
+ workspace : str
+
+ model : str
+ The model to use for completion. Must be one of the available models.
+
+ messages : typing.Sequence[GuardrailCheckRequestMessagesItem]
+ A list of messages comprising the conversation so far
+
+ response_format : typing.Optional[typing.Dict[str, typing.Any]]
+ Format of the response. Use {'type': 'json_object'} for JSON mode or {'type': 'json_schema', 'json_schema': {...}} for structured outputs.
+
+ max_tokens : typing.Optional[int]
+ The maximum number of tokens that can be generated in the chat completion.
+
+ n : typing.Optional[int]
+ How many chat completion choices to generate for each input message.
+
+ stream : typing.Optional[bool]
+ If set, partial message deltas will be sent, like in ChatGPT.
+
+ temperature : typing.Optional[float]
+ What sampling temperature to use, between 0 and 2.
+
+ top_p : typing.Optional[float]
+ An alternative to sampling with temperature, called nucleus sampling.
+
+ stop : typing.Optional[GuardrailCheckRequestStop]
+ Up to 4 sequences where the API will stop generating further tokens.
+
+ frequency_penalty : typing.Optional[float]
+ Positive values penalize new tokens based on their existing frequency in the text.
+
+ presence_penalty : typing.Optional[float]
+ Positive values penalize new tokens based on whether they appear in the text so far.
+
+ function_call : typing.Optional[GuardrailCheckRequestFunctionCall]
+ Deprecated in favor of tool_choice. 'none' means the model will not call a function and instead generates a message. 'auto' means the model can pick between generating a message or calling a function. Specifying a particular function via {'name': 'my_function'} forces the model to call that function.
+
+ seed : typing.Optional[int]
+ If specified, attempts to sample deterministically.
+
+ logit_bias : typing.Optional[typing.Dict[str, float]]
+ Modify the likelihood of specified tokens appearing in the completion. Maps token IDs (as strings) to bias values from -100 to 100.
+
+ top_logprobs : typing.Optional[int]
+ The number of most likely tokens to return at each token position.
+
+ logprobs : typing.Optional[bool]
+ Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message
+
+ tool_choice : typing.Optional[GuardrailCheckRequestToolChoice]
+ Controls which (if any) tool is called by the model. 'none' means no tool is called, 'auto' lets the model decide, 'required' forces a tool call.
+
+ user : typing.Optional[str]
+ A unique identifier representing your end-user, used by some providers for abuse monitoring.
+
+ tools : typing.Optional[typing.Sequence[typing.Dict[str, typing.Any]]]
+ A list of tools the model may call. Each tool is an object with a 'type' field and a 'function' definition.
+
+ ignore_eos : typing.Optional[bool]
+ Ignore the eos when running
+
+ reasoning_effort : typing.Optional[str]
+ Constrains effort on reasoning for reasoning models. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.
+
+ max_completion_tokens : typing.Optional[int]
+ An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens. Preferred over max_tokens for reasoning models.
+
+ stream_options : typing.Optional[typing.Dict[str, bool]]
+ Options for streaming response. Only set this when stream=True. Supports include_usage to receive token usage in the final stream chunk.
+
+ vision : typing.Optional[bool]
+ Whether this is a vision-capable request with image inputs.
+
+ guardrails : typing.Optional[GuardrailsDataInput]
+ Guardrails specific options for the request.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[GuardrailCheckResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/guardrails/v2/workspaces/{encode_path_param(workspace)}/checks",
+ method="POST",
+ json={
+ "model": model,
+ "response_format": response_format,
+ "max_tokens": max_tokens,
+ "n": n,
+ "stream": stream,
+ "temperature": temperature,
+ "top_p": top_p,
+ "stop": convert_and_respect_annotation_metadata(
+ object_=stop, annotation=GuardrailCheckRequestStop, direction="write"
+ ),
+ "frequency_penalty": frequency_penalty,
+ "presence_penalty": presence_penalty,
+ "function_call": convert_and_respect_annotation_metadata(
+ object_=function_call, annotation=GuardrailCheckRequestFunctionCall, direction="write"
+ ),
+ "seed": seed,
+ "logit_bias": logit_bias,
+ "top_logprobs": top_logprobs,
+ "logprobs": logprobs,
+ "tool_choice": convert_and_respect_annotation_metadata(
+ object_=tool_choice, annotation=GuardrailCheckRequestToolChoice, direction="write"
+ ),
+ "user": user,
+ "tools": tools,
+ "ignore_eos": ignore_eos,
+ "reasoning_effort": reasoning_effort,
+ "max_completion_tokens": max_completion_tokens,
+ "stream_options": stream_options,
+ "messages": convert_and_respect_annotation_metadata(
+ object_=messages, annotation=typing.Sequence[GuardrailCheckRequestMessagesItem], direction="write"
+ ),
+ "vision": vision,
+ "guardrails": convert_and_respect_annotation_metadata(
+ object_=guardrails, annotation=GuardrailsDataInput, direction="write"
+ ),
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ GuardrailCheckResponse,
+ parse_obj_as(
+ type_=GuardrailCheckResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 400:
+ raise BadRequestError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 500:
+ raise InternalServerError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def list_guardrail_configs(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[GenericSortField] = None,
+ filter: typing.Optional[GuardrailConfigFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[GuardrailConfigsPage]:
+ """
+ List available guardrail configs.
+
+ Lists guardrail configs for a specific workspace.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[GenericSortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[GuardrailConfigFilter]
+ Filter guardrail configs by name, description, project, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[GuardrailConfigsPage]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/guardrails/v2/workspaces/{encode_path_param(workspace)}/configs",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=GuardrailConfigFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ GuardrailConfigsPage,
+ parse_obj_as(
+ type_=GuardrailConfigsPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def create_config(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ description: typing.Optional[str] = OMIT,
+ data: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[GuardrailConfig]:
+ """
+ Create a new guardrail config.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ The name of the guardrail config
+
+ description : typing.Optional[str]
+ Description of the guardrail config
+
+ data : typing.Optional[typing.Dict[str, typing.Any]]
+ Guardrail configuration data
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[GuardrailConfig]
+ Config created successfully.
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/guardrails/v2/workspaces/{encode_path_param(workspace)}/configs",
+ method="POST",
+ json={
+ "name": name,
+ "description": description,
+ "data": data,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ GuardrailConfig,
+ parse_obj_as(
+ type_=GuardrailConfig, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 409:
+ raise ConflictError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_guardrail_config(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[GuardrailConfig]:
+ """
+ Get info about a guardrail configuration.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[GuardrailConfig]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/guardrails/v2/workspaces/{encode_path_param(workspace)}/configs/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ GuardrailConfig,
+ parse_obj_as(
+ type_=GuardrailConfig, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def delete_config(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[DeleteResponse]:
+ """
+ Delete a guardrail config.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[DeleteResponse]
+ Successful model deletion.
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/guardrails/v2/workspaces/{encode_path_param(workspace)}/configs/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ DeleteResponse,
+ parse_obj_as(
+ type_=DeleteResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def update_config(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ data: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[GuardrailConfig]:
+ """
+ Update model metadata. If the request body has an empty field,
+ keep the old value.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ description : typing.Optional[str]
+ Description of the guardrail config
+
+ data : typing.Optional[typing.Dict[str, typing.Any]]
+ Guardrail configuration data
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[GuardrailConfig]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/guardrails/v2/workspaces/{encode_path_param(workspace)}/configs/{encode_path_param(name)}",
+ method="PATCH",
+ json={
+ "description": description,
+ "data": data,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ GuardrailConfig,
+ parse_obj_as(
+ type_=GuardrailConfig, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
diff --git a/sdks/python/guardrails/types/__init__.py b/sdks/python/guardrails/types/__init__.py
new file mode 100644
index 0000000000..37092c0fd6
--- /dev/null
+++ b/sdks/python/guardrails/types/__init__.py
@@ -0,0 +1,64 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
+import typing
+from importlib import import_module
+
+if typing.TYPE_CHECKING:
+ from .guardrail_check_request_function_call import GuardrailCheckRequestFunctionCall
+ from .guardrail_check_request_messages_item import (
+ GuardrailCheckRequestMessagesItem,
+ GuardrailCheckRequestMessagesItem_Assistant,
+ GuardrailCheckRequestMessagesItem_Function,
+ GuardrailCheckRequestMessagesItem_System,
+ GuardrailCheckRequestMessagesItem_Tool,
+ GuardrailCheckRequestMessagesItem_User,
+ )
+ from .guardrail_check_request_stop import GuardrailCheckRequestStop
+ from .guardrail_check_request_tool_choice import GuardrailCheckRequestToolChoice
+_dynamic_imports: typing.Dict[str, str] = {
+ "GuardrailCheckRequestFunctionCall": ".guardrail_check_request_function_call",
+ "GuardrailCheckRequestMessagesItem": ".guardrail_check_request_messages_item",
+ "GuardrailCheckRequestMessagesItem_Assistant": ".guardrail_check_request_messages_item",
+ "GuardrailCheckRequestMessagesItem_Function": ".guardrail_check_request_messages_item",
+ "GuardrailCheckRequestMessagesItem_System": ".guardrail_check_request_messages_item",
+ "GuardrailCheckRequestMessagesItem_Tool": ".guardrail_check_request_messages_item",
+ "GuardrailCheckRequestMessagesItem_User": ".guardrail_check_request_messages_item",
+ "GuardrailCheckRequestStop": ".guardrail_check_request_stop",
+ "GuardrailCheckRequestToolChoice": ".guardrail_check_request_tool_choice",
+}
+
+
+def __getattr__(attr_name: str) -> typing.Any:
+ module_name = _dynamic_imports.get(attr_name)
+ if module_name is None:
+ raise AttributeError(f"No {attr_name} found in _dynamic_imports for module name -> {__name__}")
+ try:
+ module = import_module(module_name, __package__)
+ if module_name == f".{attr_name}":
+ return module
+ else:
+ return getattr(module, attr_name)
+ except ImportError as e:
+ raise ImportError(f"Failed to import {attr_name} from {module_name}: {e}") from e
+ except AttributeError as e:
+ raise AttributeError(f"Failed to get {attr_name} from {module_name}: {e}") from e
+
+
+def __dir__():
+ lazy_attrs = list(_dynamic_imports.keys())
+ return sorted(lazy_attrs)
+
+
+__all__ = [
+ "GuardrailCheckRequestFunctionCall",
+ "GuardrailCheckRequestMessagesItem",
+ "GuardrailCheckRequestMessagesItem_Assistant",
+ "GuardrailCheckRequestMessagesItem_Function",
+ "GuardrailCheckRequestMessagesItem_System",
+ "GuardrailCheckRequestMessagesItem_Tool",
+ "GuardrailCheckRequestMessagesItem_User",
+ "GuardrailCheckRequestStop",
+ "GuardrailCheckRequestToolChoice",
+]
diff --git a/sdks/python/guardrails/types/guardrail_check_request_function_call.py b/sdks/python/guardrails/types/guardrail_check_request_function_call.py
new file mode 100644
index 0000000000..89ac25416a
--- /dev/null
+++ b/sdks/python/guardrails/types/guardrail_check_request_function_call.py
@@ -0,0 +1,5 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+GuardrailCheckRequestFunctionCall = typing.Union[str, typing.Dict[str, typing.Any]]
diff --git a/sdks/python/guardrails/types/guardrail_check_request_messages_item.py b/sdks/python/guardrails/types/guardrail_check_request_messages_item.py
new file mode 100644
index 0000000000..a264b97e1b
--- /dev/null
+++ b/sdks/python/guardrails/types/guardrail_check_request_messages_item.py
@@ -0,0 +1,101 @@
+# This file was auto-generated by Fern from our API Definition.
+
+from __future__ import annotations
+
+import typing
+
+import pydantic
+import typing_extensions
+from ...core.pydantic_utilities import IS_PYDANTIC_V2, UniversalBaseModel
+from ...types.chat_completion_message_tool_call_param import ChatCompletionMessageToolCallParam
+from ...types.chat_completion_user_message_param_content import ChatCompletionUserMessageParamContent
+from ...types.function_call import FunctionCall
+
+
+class GuardrailCheckRequestMessagesItem_System(UniversalBaseModel):
+ role: typing.Literal["system"] = "system"
+ content: str
+ name: typing.Optional[str] = None
+
+ if IS_PYDANTIC_V2:
+ model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True) # type: ignore # Pydantic v2
+ else:
+
+ class Config:
+ frozen = True
+ smart_union = True
+ extra = pydantic.Extra.allow
+
+
+class GuardrailCheckRequestMessagesItem_User(UniversalBaseModel):
+ role: typing.Literal["user"] = "user"
+ content: ChatCompletionUserMessageParamContent
+ name: typing.Optional[str] = None
+
+ if IS_PYDANTIC_V2:
+ model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True) # type: ignore # Pydantic v2
+ else:
+
+ class Config:
+ frozen = True
+ smart_union = True
+ extra = pydantic.Extra.allow
+
+
+class GuardrailCheckRequestMessagesItem_Assistant(UniversalBaseModel):
+ role: typing.Literal["assistant"] = "assistant"
+ content: typing.Optional[str] = None
+ function_call: typing.Optional[FunctionCall] = None
+ name: typing.Optional[str] = None
+ tool_calls: typing.Optional[typing.List[ChatCompletionMessageToolCallParam]] = None
+
+ if IS_PYDANTIC_V2:
+ model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True) # type: ignore # Pydantic v2
+ else:
+
+ class Config:
+ frozen = True
+ smart_union = True
+ extra = pydantic.Extra.allow
+
+
+class GuardrailCheckRequestMessagesItem_Tool(UniversalBaseModel):
+ role: typing.Literal["tool"] = "tool"
+ content: str
+ tool_call_id: str
+
+ if IS_PYDANTIC_V2:
+ model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True) # type: ignore # Pydantic v2
+ else:
+
+ class Config:
+ frozen = True
+ smart_union = True
+ extra = pydantic.Extra.allow
+
+
+class GuardrailCheckRequestMessagesItem_Function(UniversalBaseModel):
+ role: typing.Literal["function"] = "function"
+ content: str
+ name: str
+
+ if IS_PYDANTIC_V2:
+ model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True) # type: ignore # Pydantic v2
+ else:
+
+ class Config:
+ frozen = True
+ smart_union = True
+ extra = pydantic.Extra.allow
+
+
+GuardrailCheckRequestMessagesItem = typing_extensions.Annotated[
+ typing.Union[
+ GuardrailCheckRequestMessagesItem_System,
+ GuardrailCheckRequestMessagesItem_User,
+ GuardrailCheckRequestMessagesItem_Assistant,
+ GuardrailCheckRequestMessagesItem_Tool,
+ GuardrailCheckRequestMessagesItem_Function,
+ ],
+ pydantic.Field(discriminator="role"),
+]
diff --git a/sdks/python/guardrails/types/guardrail_check_request_stop.py b/sdks/python/guardrails/types/guardrail_check_request_stop.py
new file mode 100644
index 0000000000..70abcaeef0
--- /dev/null
+++ b/sdks/python/guardrails/types/guardrail_check_request_stop.py
@@ -0,0 +1,5 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+GuardrailCheckRequestStop = typing.Union[str, typing.List[str]]
diff --git a/sdks/python/guardrails/types/guardrail_check_request_tool_choice.py b/sdks/python/guardrails/types/guardrail_check_request_tool_choice.py
new file mode 100644
index 0000000000..66dc40329b
--- /dev/null
+++ b/sdks/python/guardrails/types/guardrail_check_request_tool_choice.py
@@ -0,0 +1,5 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+GuardrailCheckRequestToolChoice = typing.Union[str, typing.Dict[str, typing.Any]]
diff --git a/sdks/python/iam/__init__.py b/sdks/python/iam/__init__.py
new file mode 100644
index 0000000000..5cde0202dc
--- /dev/null
+++ b/sdks/python/iam/__init__.py
@@ -0,0 +1,4 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
diff --git a/sdks/python/iam/client.py b/sdks/python/iam/client.py
new file mode 100644
index 0000000000..e61947a887
--- /dev/null
+++ b/sdks/python/iam/client.py
@@ -0,0 +1,428 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.request_options import RequestOptions
+from ..types.delete_response import DeleteResponse
+from ..types.role_binding import RoleBinding
+from ..types.role_binding_filter import RoleBindingFilter
+from ..types.role_bindings_page import RoleBindingsPage
+from .raw_client import AsyncRawIamClient, RawIamClient
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class IamClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._raw_client = RawIamClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> RawIamClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ RawIamClient
+ """
+ return self._raw_client
+
+ def list_role_bindings(
+ self,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[str] = None,
+ filter: typing.Optional[RoleBindingFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> RoleBindingsPage:
+ """
+ List all role bindings (Platform Admin only)
+
+ Parameters
+ ----------
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[str]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[RoleBindingFilter]
+ Filter role bindings by principal, workspace, role, granted_by, is_active, granted_at, and revoked_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ RoleBindingsPage
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.iam.list_role_bindings()
+ """
+ _response = self._raw_client.list_role_bindings(
+ page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ def create_role_binding(
+ self,
+ *,
+ principal: str,
+ role: str,
+ wait_role_propagation: typing.Optional[bool] = None,
+ workspace: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> RoleBinding:
+ """
+ Create a new role binding (Platform Admin only)
+
+ Parameters
+ ----------
+ principal : str
+ The principal identifier (email, user ID, or group ID)
+
+ role : str
+ The role name (e.g., 'Viewer', 'Editor', 'Admin')
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for role to propagate before returning (default: true). Set to false for bulk operations.
+
+ workspace : typing.Optional[str]
+ The workspace this binding applies to. None for platform-level roles.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ RoleBinding
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.iam.create_role_binding(
+ principal="principal",
+ role="role",
+ )
+ """
+ _response = self._raw_client.create_role_binding(
+ principal=principal,
+ role=role,
+ wait_role_propagation=wait_role_propagation,
+ workspace=workspace,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def get_role_binding(self, name: str, *, request_options: typing.Optional[RequestOptions] = None) -> RoleBinding:
+ """
+ Get a specific role binding (Platform Admin only)
+
+ Parameters
+ ----------
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ RoleBinding
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.iam.get_role_binding(
+ name="name",
+ )
+ """
+ _response = self._raw_client.get_role_binding(name, request_options=request_options)
+ return _response.data
+
+ def revoke_role_binding(
+ self,
+ name: str,
+ *,
+ wait_role_propagation: typing.Optional[bool] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> DeleteResponse:
+ """
+ Revoke a role binding (Platform Admin only)
+
+ Parameters
+ ----------
+ name : str
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for role to propagate before returning (default: true). Set to false for bulk operations.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ DeleteResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.iam.revoke_role_binding(
+ name="name",
+ )
+ """
+ _response = self._raw_client.revoke_role_binding(
+ name, wait_role_propagation=wait_role_propagation, request_options=request_options
+ )
+ return _response.data
+
+
+class AsyncIamClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._raw_client = AsyncRawIamClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> AsyncRawIamClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ AsyncRawIamClient
+ """
+ return self._raw_client
+
+ async def list_role_bindings(
+ self,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[str] = None,
+ filter: typing.Optional[RoleBindingFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> RoleBindingsPage:
+ """
+ List all role bindings (Platform Admin only)
+
+ Parameters
+ ----------
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[str]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[RoleBindingFilter]
+ Filter role bindings by principal, workspace, role, granted_by, is_active, granted_at, and revoked_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ RoleBindingsPage
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.iam.list_role_bindings()
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_role_bindings(
+ page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ async def create_role_binding(
+ self,
+ *,
+ principal: str,
+ role: str,
+ wait_role_propagation: typing.Optional[bool] = None,
+ workspace: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> RoleBinding:
+ """
+ Create a new role binding (Platform Admin only)
+
+ Parameters
+ ----------
+ principal : str
+ The principal identifier (email, user ID, or group ID)
+
+ role : str
+ The role name (e.g., 'Viewer', 'Editor', 'Admin')
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for role to propagate before returning (default: true). Set to false for bulk operations.
+
+ workspace : typing.Optional[str]
+ The workspace this binding applies to. None for platform-level roles.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ RoleBinding
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.iam.create_role_binding(
+ principal="principal",
+ role="role",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.create_role_binding(
+ principal=principal,
+ role=role,
+ wait_role_propagation=wait_role_propagation,
+ workspace=workspace,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def get_role_binding(
+ self, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> RoleBinding:
+ """
+ Get a specific role binding (Platform Admin only)
+
+ Parameters
+ ----------
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ RoleBinding
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.iam.get_role_binding(
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_role_binding(name, request_options=request_options)
+ return _response.data
+
+ async def revoke_role_binding(
+ self,
+ name: str,
+ *,
+ wait_role_propagation: typing.Optional[bool] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> DeleteResponse:
+ """
+ Revoke a role binding (Platform Admin only)
+
+ Parameters
+ ----------
+ name : str
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for role to propagate before returning (default: true). Set to false for bulk operations.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ DeleteResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.iam.revoke_role_binding(
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.revoke_role_binding(
+ name, wait_role_propagation=wait_role_propagation, request_options=request_options
+ )
+ return _response.data
diff --git a/sdks/python/iam/raw_client.py b/sdks/python/iam/raw_client.py
new file mode 100644
index 0000000000..57e8d69144
--- /dev/null
+++ b/sdks/python/iam/raw_client.py
@@ -0,0 +1,580 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+from json.decoder import JSONDecodeError
+
+from ..core.api_error import ApiError
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.http_response import AsyncHttpResponse, HttpResponse
+from ..core.jsonable_encoder import encode_path_param
+from ..core.parse_error import ParsingError
+from ..core.pydantic_utilities import parse_obj_as
+from ..core.request_options import RequestOptions
+from ..core.serialization import convert_and_respect_annotation_metadata
+from ..errors.unprocessable_entity_error import UnprocessableEntityError
+from ..types.delete_response import DeleteResponse
+from ..types.role_binding import RoleBinding
+from ..types.role_binding_filter import RoleBindingFilter
+from ..types.role_bindings_page import RoleBindingsPage
+from pydantic import ValidationError
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class RawIamClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ def list_role_bindings(
+ self,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[str] = None,
+ filter: typing.Optional[RoleBindingFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[RoleBindingsPage]:
+ """
+ List all role bindings (Platform Admin only)
+
+ Parameters
+ ----------
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[str]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[RoleBindingFilter]
+ Filter role bindings by principal, workspace, role, granted_by, is_active, granted_at, and revoked_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[RoleBindingsPage]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ "apis/auth/v2/iam/role-bindings",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=RoleBindingFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ RoleBindingsPage,
+ parse_obj_as(
+ type_=RoleBindingsPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def create_role_binding(
+ self,
+ *,
+ principal: str,
+ role: str,
+ wait_role_propagation: typing.Optional[bool] = None,
+ workspace: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[RoleBinding]:
+ """
+ Create a new role binding (Platform Admin only)
+
+ Parameters
+ ----------
+ principal : str
+ The principal identifier (email, user ID, or group ID)
+
+ role : str
+ The role name (e.g., 'Viewer', 'Editor', 'Admin')
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for role to propagate before returning (default: true). Set to false for bulk operations.
+
+ workspace : typing.Optional[str]
+ The workspace this binding applies to. None for platform-level roles.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[RoleBinding]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ "apis/auth/v2/iam/role-bindings",
+ method="POST",
+ params={
+ "wait_role_propagation": wait_role_propagation,
+ },
+ json={
+ "principal": principal,
+ "workspace": workspace,
+ "role": role,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ RoleBinding,
+ parse_obj_as(
+ type_=RoleBinding, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_role_binding(
+ self, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[RoleBinding]:
+ """
+ Get a specific role binding (Platform Admin only)
+
+ Parameters
+ ----------
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[RoleBinding]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/auth/v2/iam/role-bindings/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ RoleBinding,
+ parse_obj_as(
+ type_=RoleBinding, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def revoke_role_binding(
+ self,
+ name: str,
+ *,
+ wait_role_propagation: typing.Optional[bool] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[DeleteResponse]:
+ """
+ Revoke a role binding (Platform Admin only)
+
+ Parameters
+ ----------
+ name : str
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for role to propagate before returning (default: true). Set to false for bulk operations.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[DeleteResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/auth/v2/iam/role-bindings/{encode_path_param(name)}",
+ method="DELETE",
+ params={
+ "wait_role_propagation": wait_role_propagation,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ DeleteResponse,
+ parse_obj_as(
+ type_=DeleteResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+
+class AsyncRawIamClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ async def list_role_bindings(
+ self,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[str] = None,
+ filter: typing.Optional[RoleBindingFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[RoleBindingsPage]:
+ """
+ List all role bindings (Platform Admin only)
+
+ Parameters
+ ----------
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[str]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[RoleBindingFilter]
+ Filter role bindings by principal, workspace, role, granted_by, is_active, granted_at, and revoked_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[RoleBindingsPage]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ "apis/auth/v2/iam/role-bindings",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=RoleBindingFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ RoleBindingsPage,
+ parse_obj_as(
+ type_=RoleBindingsPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def create_role_binding(
+ self,
+ *,
+ principal: str,
+ role: str,
+ wait_role_propagation: typing.Optional[bool] = None,
+ workspace: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[RoleBinding]:
+ """
+ Create a new role binding (Platform Admin only)
+
+ Parameters
+ ----------
+ principal : str
+ The principal identifier (email, user ID, or group ID)
+
+ role : str
+ The role name (e.g., 'Viewer', 'Editor', 'Admin')
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for role to propagate before returning (default: true). Set to false for bulk operations.
+
+ workspace : typing.Optional[str]
+ The workspace this binding applies to. None for platform-level roles.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[RoleBinding]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ "apis/auth/v2/iam/role-bindings",
+ method="POST",
+ params={
+ "wait_role_propagation": wait_role_propagation,
+ },
+ json={
+ "principal": principal,
+ "workspace": workspace,
+ "role": role,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ RoleBinding,
+ parse_obj_as(
+ type_=RoleBinding, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_role_binding(
+ self, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[RoleBinding]:
+ """
+ Get a specific role binding (Platform Admin only)
+
+ Parameters
+ ----------
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[RoleBinding]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/auth/v2/iam/role-bindings/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ RoleBinding,
+ parse_obj_as(
+ type_=RoleBinding, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def revoke_role_binding(
+ self,
+ name: str,
+ *,
+ wait_role_propagation: typing.Optional[bool] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[DeleteResponse]:
+ """
+ Revoke a role binding (Platform Admin only)
+
+ Parameters
+ ----------
+ name : str
+
+ wait_role_propagation : typing.Optional[bool]
+ If true, wait for role to propagate before returning (default: true). Set to false for bulk operations.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[DeleteResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/auth/v2/iam/role-bindings/{encode_path_param(name)}",
+ method="DELETE",
+ params={
+ "wait_role_propagation": wait_role_propagation,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ DeleteResponse,
+ parse_obj_as(
+ type_=DeleteResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
diff --git a/sdks/python/inference_gateway/__init__.py b/sdks/python/inference_gateway/__init__.py
new file mode 100644
index 0000000000..5cde0202dc
--- /dev/null
+++ b/sdks/python/inference_gateway/__init__.py
@@ -0,0 +1,4 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
diff --git a/sdks/python/inference_gateway/client.py b/sdks/python/inference_gateway/client.py
new file mode 100644
index 0000000000..7d83360a00
--- /dev/null
+++ b/sdks/python/inference_gateway/client.py
@@ -0,0 +1,1890 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.request_options import RequestOptions
+from ..types.open_ai_list_models_resp import OpenAiListModelsResp
+from ..types.open_ai_model_resp import OpenAiModelResp
+from .raw_client import AsyncRawInferenceGatewayClient, RawInferenceGatewayClient
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class InferenceGatewayClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._raw_client = RawInferenceGatewayClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> RawInferenceGatewayClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ RawInferenceGatewayClient
+ """
+ return self._raw_client
+
+ def gateway_proxy_get(
+ self, workspace: str, name: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Any:
+ """
+ Proxy requests to model entity inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Any
+ Proxy GET request to model entity inference endpoint
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.inference_gateway.gateway_proxy_get(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ )
+ """
+ _response = self._raw_client.gateway_proxy_get(workspace, name, trailing_uri, request_options=request_options)
+ return _response.data
+
+ def gateway_proxy_post(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Proxy requests to model entity inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Proxy POST request to model entity inference endpoint
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.inference_gateway.gateway_proxy_post(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ request={"key": "value"},
+ )
+ """
+ _response = self._raw_client.gateway_proxy_post(
+ workspace, name, trailing_uri, request=request, request_options=request_options
+ )
+ return _response.data
+
+ def gateway_proxy_put(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Proxy requests to model entity inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Proxy PUT request to model entity inference endpoint
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.inference_gateway.gateway_proxy_put(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ request={"key": "value"},
+ )
+ """
+ _response = self._raw_client.gateway_proxy_put(
+ workspace, name, trailing_uri, request=request, request_options=request_options
+ )
+ return _response.data
+
+ def gateway_proxy_delete(
+ self, workspace: str, name: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Any:
+ """
+ Proxy requests to model entity inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Any
+ Proxy DELETE request to model entity inference endpoint
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.inference_gateway.gateway_proxy_delete(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ )
+ """
+ _response = self._raw_client.gateway_proxy_delete(
+ workspace, name, trailing_uri, request_options=request_options
+ )
+ return _response.data
+
+ def gateway_proxy_patch(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Proxy requests to model entity inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Proxy PATCH request to model entity inference endpoint
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.inference_gateway.gateway_proxy_patch(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ request={"key": "value"},
+ )
+ """
+ _response = self._raw_client.gateway_proxy_patch(
+ workspace, name, trailing_uri, request=request, request_options=request_options
+ )
+ return _response.data
+
+ def openai_proxy_list_models(
+ self, workspace: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> OpenAiListModelsResp:
+ """
+ This endpoint aggregates models from all model entities and returns them
+ in OpenAI's list models format. Each model ID is the model entity identifier
+ in format workspace/model_entity_name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ OpenAiListModelsResp
+ List models request to OpenAI-compatible endpoint
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.inference_gateway.openai_proxy_list_models(
+ workspace="workspace",
+ )
+ """
+ _response = self._raw_client.openai_proxy_list_models(workspace, request_options=request_options)
+ return _response.data
+
+ def openai_proxy_get_model(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> OpenAiModelResp:
+ """
+ Retrieve information about a specific OpenAI-compatible model.
+ Workspace is always taken from the URL path; name may be model_entity_name
+ or workspace/model_entity_name (workspace prefix is ignored).
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ OpenAiModelResp
+ Get model request to OpenAI-compatible endpoint
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.inference_gateway.openai_proxy_get_model(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.openai_proxy_get_model(workspace, name, request_options=request_options)
+ return _response.data
+
+ def openai_proxy_get(
+ self, workspace: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Any:
+ """
+ Proxy requests to OpenAI-compatible inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Any
+ Proxy GET request to OpenAI-compatible endpoint
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.inference_gateway.openai_proxy_get(
+ workspace="workspace",
+ trailing_uri="trailing_uri",
+ )
+ """
+ _response = self._raw_client.openai_proxy_get(workspace, trailing_uri, request_options=request_options)
+ return _response.data
+
+ def openai_proxy_post(
+ self,
+ workspace: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Proxy requests to OpenAI-compatible inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Proxy POST request to OpenAI-compatible endpoint
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.inference_gateway.openai_proxy_post(
+ workspace="workspace",
+ trailing_uri="trailing_uri",
+ request={"key": "value"},
+ )
+ """
+ _response = self._raw_client.openai_proxy_post(
+ workspace, trailing_uri, request=request, request_options=request_options
+ )
+ return _response.data
+
+ def openai_proxy_put(
+ self,
+ workspace: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Proxy requests to OpenAI-compatible inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Proxy PUT request to OpenAI-compatible endpoint
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.inference_gateway.openai_proxy_put(
+ workspace="workspace",
+ trailing_uri="trailing_uri",
+ request={"key": "value"},
+ )
+ """
+ _response = self._raw_client.openai_proxy_put(
+ workspace, trailing_uri, request=request, request_options=request_options
+ )
+ return _response.data
+
+ def openai_proxy_delete(
+ self, workspace: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Any:
+ """
+ Proxy requests to OpenAI-compatible inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Any
+ Proxy DELETE request to OpenAI-compatible endpoint
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.inference_gateway.openai_proxy_delete(
+ workspace="workspace",
+ trailing_uri="trailing_uri",
+ )
+ """
+ _response = self._raw_client.openai_proxy_delete(workspace, trailing_uri, request_options=request_options)
+ return _response.data
+
+ def openai_proxy_patch(
+ self,
+ workspace: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Proxy requests to OpenAI-compatible inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Proxy PATCH request to OpenAI-compatible endpoint
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.inference_gateway.openai_proxy_patch(
+ workspace="workspace",
+ trailing_uri="trailing_uri",
+ request={"key": "value"},
+ )
+ """
+ _response = self._raw_client.openai_proxy_patch(
+ workspace, trailing_uri, request=request, request_options=request_options
+ )
+ return _response.data
+
+ def provider_proxy_get(
+ self, workspace: str, name: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Any:
+ """
+ Proxy requests to provider inference endpoints.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Any
+ Proxy GET request to provider inference endpoint
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.inference_gateway.provider_proxy_get(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ )
+ """
+ _response = self._raw_client.provider_proxy_get(workspace, name, trailing_uri, request_options=request_options)
+ return _response.data
+
+ def provider_proxy_post(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Proxy requests to provider inference endpoints.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Proxy POST request to provider inference endpoint
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.inference_gateway.provider_proxy_post(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ request={"key": "value"},
+ )
+ """
+ _response = self._raw_client.provider_proxy_post(
+ workspace, name, trailing_uri, request=request, request_options=request_options
+ )
+ return _response.data
+
+ def provider_proxy_put(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Proxy requests to provider inference endpoints.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Proxy PUT request to provider inference endpoint
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.inference_gateway.provider_proxy_put(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ request={"key": "value"},
+ )
+ """
+ _response = self._raw_client.provider_proxy_put(
+ workspace, name, trailing_uri, request=request, request_options=request_options
+ )
+ return _response.data
+
+ def provider_proxy_delete(
+ self, workspace: str, name: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Any:
+ """
+ Proxy requests to provider inference endpoints.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Any
+ Proxy DELETE request to provider inference endpoint
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.inference_gateway.provider_proxy_delete(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ )
+ """
+ _response = self._raw_client.provider_proxy_delete(
+ workspace, name, trailing_uri, request_options=request_options
+ )
+ return _response.data
+
+ def provider_proxy_patch(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Proxy requests to provider inference endpoints.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Proxy PATCH request to provider inference endpoint
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.inference_gateway.provider_proxy_patch(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ request={"key": "value"},
+ )
+ """
+ _response = self._raw_client.provider_proxy_patch(
+ workspace, name, trailing_uri, request=request, request_options=request_options
+ )
+ return _response.data
+
+ def provider_ready(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Check if a model provider is registered in the gateway's cache.
+
+ This is a lightweight endpoint that only checks the gateway's internal state,
+ without making any requests to the actual provider backend. Use this to verify
+ the gateway is ready to route requests to a provider after deployment.
+
+ Returns:
+ 200 OK with provider info if the provider is registered
+ 404 Not Found if the provider is not yet in the gateway's cache
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Check if the gateway can route to a provider
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.inference_gateway.provider_ready(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.provider_ready(workspace, name, request_options=request_options)
+ return _response.data
+
+
+class AsyncInferenceGatewayClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._raw_client = AsyncRawInferenceGatewayClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> AsyncRawInferenceGatewayClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ AsyncRawInferenceGatewayClient
+ """
+ return self._raw_client
+
+ async def gateway_proxy_get(
+ self, workspace: str, name: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Any:
+ """
+ Proxy requests to model entity inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Any
+ Proxy GET request to model entity inference endpoint
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.inference_gateway.gateway_proxy_get(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.gateway_proxy_get(
+ workspace, name, trailing_uri, request_options=request_options
+ )
+ return _response.data
+
+ async def gateway_proxy_post(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Proxy requests to model entity inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Proxy POST request to model entity inference endpoint
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.inference_gateway.gateway_proxy_post(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ request={"key": "value"},
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.gateway_proxy_post(
+ workspace, name, trailing_uri, request=request, request_options=request_options
+ )
+ return _response.data
+
+ async def gateway_proxy_put(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Proxy requests to model entity inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Proxy PUT request to model entity inference endpoint
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.inference_gateway.gateway_proxy_put(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ request={"key": "value"},
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.gateway_proxy_put(
+ workspace, name, trailing_uri, request=request, request_options=request_options
+ )
+ return _response.data
+
+ async def gateway_proxy_delete(
+ self, workspace: str, name: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Any:
+ """
+ Proxy requests to model entity inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Any
+ Proxy DELETE request to model entity inference endpoint
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.inference_gateway.gateway_proxy_delete(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.gateway_proxy_delete(
+ workspace, name, trailing_uri, request_options=request_options
+ )
+ return _response.data
+
+ async def gateway_proxy_patch(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Proxy requests to model entity inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Proxy PATCH request to model entity inference endpoint
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.inference_gateway.gateway_proxy_patch(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ request={"key": "value"},
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.gateway_proxy_patch(
+ workspace, name, trailing_uri, request=request, request_options=request_options
+ )
+ return _response.data
+
+ async def openai_proxy_list_models(
+ self, workspace: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> OpenAiListModelsResp:
+ """
+ This endpoint aggregates models from all model entities and returns them
+ in OpenAI's list models format. Each model ID is the model entity identifier
+ in format workspace/model_entity_name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ OpenAiListModelsResp
+ List models request to OpenAI-compatible endpoint
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.inference_gateway.openai_proxy_list_models(
+ workspace="workspace",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.openai_proxy_list_models(workspace, request_options=request_options)
+ return _response.data
+
+ async def openai_proxy_get_model(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> OpenAiModelResp:
+ """
+ Retrieve information about a specific OpenAI-compatible model.
+ Workspace is always taken from the URL path; name may be model_entity_name
+ or workspace/model_entity_name (workspace prefix is ignored).
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ OpenAiModelResp
+ Get model request to OpenAI-compatible endpoint
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.inference_gateway.openai_proxy_get_model(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.openai_proxy_get_model(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def openai_proxy_get(
+ self, workspace: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Any:
+ """
+ Proxy requests to OpenAI-compatible inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Any
+ Proxy GET request to OpenAI-compatible endpoint
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.inference_gateway.openai_proxy_get(
+ workspace="workspace",
+ trailing_uri="trailing_uri",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.openai_proxy_get(workspace, trailing_uri, request_options=request_options)
+ return _response.data
+
+ async def openai_proxy_post(
+ self,
+ workspace: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Proxy requests to OpenAI-compatible inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Proxy POST request to OpenAI-compatible endpoint
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.inference_gateway.openai_proxy_post(
+ workspace="workspace",
+ trailing_uri="trailing_uri",
+ request={"key": "value"},
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.openai_proxy_post(
+ workspace, trailing_uri, request=request, request_options=request_options
+ )
+ return _response.data
+
+ async def openai_proxy_put(
+ self,
+ workspace: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Proxy requests to OpenAI-compatible inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Proxy PUT request to OpenAI-compatible endpoint
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.inference_gateway.openai_proxy_put(
+ workspace="workspace",
+ trailing_uri="trailing_uri",
+ request={"key": "value"},
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.openai_proxy_put(
+ workspace, trailing_uri, request=request, request_options=request_options
+ )
+ return _response.data
+
+ async def openai_proxy_delete(
+ self, workspace: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Any:
+ """
+ Proxy requests to OpenAI-compatible inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Any
+ Proxy DELETE request to OpenAI-compatible endpoint
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.inference_gateway.openai_proxy_delete(
+ workspace="workspace",
+ trailing_uri="trailing_uri",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.openai_proxy_delete(workspace, trailing_uri, request_options=request_options)
+ return _response.data
+
+ async def openai_proxy_patch(
+ self,
+ workspace: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Proxy requests to OpenAI-compatible inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Proxy PATCH request to OpenAI-compatible endpoint
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.inference_gateway.openai_proxy_patch(
+ workspace="workspace",
+ trailing_uri="trailing_uri",
+ request={"key": "value"},
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.openai_proxy_patch(
+ workspace, trailing_uri, request=request, request_options=request_options
+ )
+ return _response.data
+
+ async def provider_proxy_get(
+ self, workspace: str, name: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Any:
+ """
+ Proxy requests to provider inference endpoints.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Any
+ Proxy GET request to provider inference endpoint
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.inference_gateway.provider_proxy_get(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.provider_proxy_get(
+ workspace, name, trailing_uri, request_options=request_options
+ )
+ return _response.data
+
+ async def provider_proxy_post(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Proxy requests to provider inference endpoints.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Proxy POST request to provider inference endpoint
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.inference_gateway.provider_proxy_post(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ request={"key": "value"},
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.provider_proxy_post(
+ workspace, name, trailing_uri, request=request, request_options=request_options
+ )
+ return _response.data
+
+ async def provider_proxy_put(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Proxy requests to provider inference endpoints.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Proxy PUT request to provider inference endpoint
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.inference_gateway.provider_proxy_put(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ request={"key": "value"},
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.provider_proxy_put(
+ workspace, name, trailing_uri, request=request, request_options=request_options
+ )
+ return _response.data
+
+ async def provider_proxy_delete(
+ self, workspace: str, name: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Any:
+ """
+ Proxy requests to provider inference endpoints.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Any
+ Proxy DELETE request to provider inference endpoint
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.inference_gateway.provider_proxy_delete(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.provider_proxy_delete(
+ workspace, name, trailing_uri, request_options=request_options
+ )
+ return _response.data
+
+ async def provider_proxy_patch(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Proxy requests to provider inference endpoints.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Proxy PATCH request to provider inference endpoint
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.inference_gateway.provider_proxy_patch(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ request={"key": "value"},
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.provider_proxy_patch(
+ workspace, name, trailing_uri, request=request, request_options=request_options
+ )
+ return _response.data
+
+ async def provider_ready(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Check if a model provider is registered in the gateway's cache.
+
+ This is a lightweight endpoint that only checks the gateway's internal state,
+ without making any requests to the actual provider backend. Use this to verify
+ the gateway is ready to route requests to a provider after deployment.
+
+ Returns:
+ 200 OK with provider info if the provider is registered
+ 404 Not Found if the provider is not yet in the gateway's cache
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Check if the gateway can route to a provider
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.inference_gateway.provider_ready(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.provider_ready(workspace, name, request_options=request_options)
+ return _response.data
diff --git a/sdks/python/inference_gateway/raw_client.py b/sdks/python/inference_gateway/raw_client.py
new file mode 100644
index 0000000000..a064be77b0
--- /dev/null
+++ b/sdks/python/inference_gateway/raw_client.py
@@ -0,0 +1,2481 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+from json.decoder import JSONDecodeError
+
+from ..core.api_error import ApiError
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.http_response import AsyncHttpResponse, HttpResponse
+from ..core.jsonable_encoder import encode_path_param
+from ..core.parse_error import ParsingError
+from ..core.pydantic_utilities import parse_obj_as
+from ..core.request_options import RequestOptions
+from ..errors.unprocessable_entity_error import UnprocessableEntityError
+from ..types.open_ai_list_models_resp import OpenAiListModelsResp
+from ..types.open_ai_model_resp import OpenAiModelResp
+from pydantic import ValidationError
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class RawInferenceGatewayClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ def gateway_proxy_get(
+ self, workspace: str, name: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[typing.Any]:
+ """
+ Proxy requests to model entity inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Any]
+ Proxy GET request to model entity inference endpoint
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/model/{encode_path_param(name)}/-/{encode_path_param(trailing_uri)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if _response is None or not _response.text.strip():
+ return HttpResponse(response=_response, data=None)
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def gateway_proxy_post(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Proxy requests to model entity inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Dict[str, typing.Any]]
+ Proxy POST request to model entity inference endpoint
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/model/{encode_path_param(name)}/-/{encode_path_param(trailing_uri)}",
+ method="POST",
+ json=request,
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def gateway_proxy_put(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Proxy requests to model entity inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Dict[str, typing.Any]]
+ Proxy PUT request to model entity inference endpoint
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/model/{encode_path_param(name)}/-/{encode_path_param(trailing_uri)}",
+ method="PUT",
+ json=request,
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def gateway_proxy_delete(
+ self, workspace: str, name: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[typing.Any]:
+ """
+ Proxy requests to model entity inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Any]
+ Proxy DELETE request to model entity inference endpoint
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/model/{encode_path_param(name)}/-/{encode_path_param(trailing_uri)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if _response is None or not _response.text.strip():
+ return HttpResponse(response=_response, data=None)
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def gateway_proxy_patch(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Proxy requests to model entity inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Dict[str, typing.Any]]
+ Proxy PATCH request to model entity inference endpoint
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/model/{encode_path_param(name)}/-/{encode_path_param(trailing_uri)}",
+ method="PATCH",
+ json=request,
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def openai_proxy_list_models(
+ self, workspace: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[OpenAiListModelsResp]:
+ """
+ This endpoint aggregates models from all model entities and returns them
+ in OpenAI's list models format. Each model ID is the model entity identifier
+ in format workspace/model_entity_name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[OpenAiListModelsResp]
+ List models request to OpenAI-compatible endpoint
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/openai/-/v1/models",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ OpenAiListModelsResp,
+ parse_obj_as(
+ type_=OpenAiListModelsResp, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def openai_proxy_get_model(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[OpenAiModelResp]:
+ """
+ Retrieve information about a specific OpenAI-compatible model.
+ Workspace is always taken from the URL path; name may be model_entity_name
+ or workspace/model_entity_name (workspace prefix is ignored).
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[OpenAiModelResp]
+ Get model request to OpenAI-compatible endpoint
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/openai/-/v1/models/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ OpenAiModelResp,
+ parse_obj_as(
+ type_=OpenAiModelResp, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def openai_proxy_get(
+ self, workspace: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[typing.Any]:
+ """
+ Proxy requests to OpenAI-compatible inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Any]
+ Proxy GET request to OpenAI-compatible endpoint
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/openai/-/{encode_path_param(trailing_uri)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if _response is None or not _response.text.strip():
+ return HttpResponse(response=_response, data=None)
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def openai_proxy_post(
+ self,
+ workspace: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Proxy requests to OpenAI-compatible inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Dict[str, typing.Any]]
+ Proxy POST request to OpenAI-compatible endpoint
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/openai/-/{encode_path_param(trailing_uri)}",
+ method="POST",
+ json=request,
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def openai_proxy_put(
+ self,
+ workspace: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Proxy requests to OpenAI-compatible inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Dict[str, typing.Any]]
+ Proxy PUT request to OpenAI-compatible endpoint
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/openai/-/{encode_path_param(trailing_uri)}",
+ method="PUT",
+ json=request,
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def openai_proxy_delete(
+ self, workspace: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[typing.Any]:
+ """
+ Proxy requests to OpenAI-compatible inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Any]
+ Proxy DELETE request to OpenAI-compatible endpoint
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/openai/-/{encode_path_param(trailing_uri)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if _response is None or not _response.text.strip():
+ return HttpResponse(response=_response, data=None)
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def openai_proxy_patch(
+ self,
+ workspace: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Proxy requests to OpenAI-compatible inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Dict[str, typing.Any]]
+ Proxy PATCH request to OpenAI-compatible endpoint
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/openai/-/{encode_path_param(trailing_uri)}",
+ method="PATCH",
+ json=request,
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def provider_proxy_get(
+ self, workspace: str, name: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[typing.Any]:
+ """
+ Proxy requests to provider inference endpoints.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Any]
+ Proxy GET request to provider inference endpoint
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/provider/{encode_path_param(name)}/-/{encode_path_param(trailing_uri)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if _response is None or not _response.text.strip():
+ return HttpResponse(response=_response, data=None)
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def provider_proxy_post(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Proxy requests to provider inference endpoints.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Dict[str, typing.Any]]
+ Proxy POST request to provider inference endpoint
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/provider/{encode_path_param(name)}/-/{encode_path_param(trailing_uri)}",
+ method="POST",
+ json=request,
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def provider_proxy_put(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Proxy requests to provider inference endpoints.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Dict[str, typing.Any]]
+ Proxy PUT request to provider inference endpoint
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/provider/{encode_path_param(name)}/-/{encode_path_param(trailing_uri)}",
+ method="PUT",
+ json=request,
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def provider_proxy_delete(
+ self, workspace: str, name: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[typing.Any]:
+ """
+ Proxy requests to provider inference endpoints.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Any]
+ Proxy DELETE request to provider inference endpoint
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/provider/{encode_path_param(name)}/-/{encode_path_param(trailing_uri)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if _response is None or not _response.text.strip():
+ return HttpResponse(response=_response, data=None)
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def provider_proxy_patch(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Proxy requests to provider inference endpoints.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Dict[str, typing.Any]]
+ Proxy PATCH request to provider inference endpoint
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/provider/{encode_path_param(name)}/-/{encode_path_param(trailing_uri)}",
+ method="PATCH",
+ json=request,
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def provider_ready(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Check if a model provider is registered in the gateway's cache.
+
+ This is a lightweight endpoint that only checks the gateway's internal state,
+ without making any requests to the actual provider backend. Use this to verify
+ the gateway is ready to route requests to a provider after deployment.
+
+ Returns:
+ 200 OK with provider info if the provider is registered
+ 404 Not Found if the provider is not yet in the gateway's cache
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Dict[str, typing.Any]]
+ Check if the gateway can route to a provider
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/provider/{encode_path_param(name)}/ready",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+
+class AsyncRawInferenceGatewayClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ async def gateway_proxy_get(
+ self, workspace: str, name: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[typing.Any]:
+ """
+ Proxy requests to model entity inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Any]
+ Proxy GET request to model entity inference endpoint
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/model/{encode_path_param(name)}/-/{encode_path_param(trailing_uri)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if _response is None or not _response.text.strip():
+ return AsyncHttpResponse(response=_response, data=None)
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def gateway_proxy_post(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Proxy requests to model entity inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Dict[str, typing.Any]]
+ Proxy POST request to model entity inference endpoint
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/model/{encode_path_param(name)}/-/{encode_path_param(trailing_uri)}",
+ method="POST",
+ json=request,
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def gateway_proxy_put(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Proxy requests to model entity inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Dict[str, typing.Any]]
+ Proxy PUT request to model entity inference endpoint
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/model/{encode_path_param(name)}/-/{encode_path_param(trailing_uri)}",
+ method="PUT",
+ json=request,
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def gateway_proxy_delete(
+ self, workspace: str, name: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[typing.Any]:
+ """
+ Proxy requests to model entity inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Any]
+ Proxy DELETE request to model entity inference endpoint
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/model/{encode_path_param(name)}/-/{encode_path_param(trailing_uri)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if _response is None or not _response.text.strip():
+ return AsyncHttpResponse(response=_response, data=None)
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def gateway_proxy_patch(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Proxy requests to model entity inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Dict[str, typing.Any]]
+ Proxy PATCH request to model entity inference endpoint
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/model/{encode_path_param(name)}/-/{encode_path_param(trailing_uri)}",
+ method="PATCH",
+ json=request,
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def openai_proxy_list_models(
+ self, workspace: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[OpenAiListModelsResp]:
+ """
+ This endpoint aggregates models from all model entities and returns them
+ in OpenAI's list models format. Each model ID is the model entity identifier
+ in format workspace/model_entity_name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[OpenAiListModelsResp]
+ List models request to OpenAI-compatible endpoint
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/openai/-/v1/models",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ OpenAiListModelsResp,
+ parse_obj_as(
+ type_=OpenAiListModelsResp, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def openai_proxy_get_model(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[OpenAiModelResp]:
+ """
+ Retrieve information about a specific OpenAI-compatible model.
+ Workspace is always taken from the URL path; name may be model_entity_name
+ or workspace/model_entity_name (workspace prefix is ignored).
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[OpenAiModelResp]
+ Get model request to OpenAI-compatible endpoint
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/openai/-/v1/models/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ OpenAiModelResp,
+ parse_obj_as(
+ type_=OpenAiModelResp, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def openai_proxy_get(
+ self, workspace: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[typing.Any]:
+ """
+ Proxy requests to OpenAI-compatible inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Any]
+ Proxy GET request to OpenAI-compatible endpoint
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/openai/-/{encode_path_param(trailing_uri)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if _response is None or not _response.text.strip():
+ return AsyncHttpResponse(response=_response, data=None)
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def openai_proxy_post(
+ self,
+ workspace: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Proxy requests to OpenAI-compatible inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Dict[str, typing.Any]]
+ Proxy POST request to OpenAI-compatible endpoint
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/openai/-/{encode_path_param(trailing_uri)}",
+ method="POST",
+ json=request,
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def openai_proxy_put(
+ self,
+ workspace: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Proxy requests to OpenAI-compatible inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Dict[str, typing.Any]]
+ Proxy PUT request to OpenAI-compatible endpoint
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/openai/-/{encode_path_param(trailing_uri)}",
+ method="PUT",
+ json=request,
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def openai_proxy_delete(
+ self, workspace: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[typing.Any]:
+ """
+ Proxy requests to OpenAI-compatible inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Any]
+ Proxy DELETE request to OpenAI-compatible endpoint
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/openai/-/{encode_path_param(trailing_uri)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if _response is None or not _response.text.strip():
+ return AsyncHttpResponse(response=_response, data=None)
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def openai_proxy_patch(
+ self,
+ workspace: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Proxy requests to OpenAI-compatible inference endpoints.
+
+ All inference requests must resolve to a `VirtualModel`. The platform's
+ provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+ for every served model entity (named after the entity, with
+ `default_model_entity` set to the entity ref) so this is the typical case;
+ operators can also create custom VirtualModels for routing, plugin chains,
+ LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+ return `404`.
+
+ Parameters
+ ----------
+ workspace : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Dict[str, typing.Any]]
+ Proxy PATCH request to OpenAI-compatible endpoint
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/openai/-/{encode_path_param(trailing_uri)}",
+ method="PATCH",
+ json=request,
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def provider_proxy_get(
+ self, workspace: str, name: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[typing.Any]:
+ """
+ Proxy requests to provider inference endpoints.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Any]
+ Proxy GET request to provider inference endpoint
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/provider/{encode_path_param(name)}/-/{encode_path_param(trailing_uri)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if _response is None or not _response.text.strip():
+ return AsyncHttpResponse(response=_response, data=None)
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def provider_proxy_post(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Proxy requests to provider inference endpoints.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Dict[str, typing.Any]]
+ Proxy POST request to provider inference endpoint
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/provider/{encode_path_param(name)}/-/{encode_path_param(trailing_uri)}",
+ method="POST",
+ json=request,
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def provider_proxy_put(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Proxy requests to provider inference endpoints.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Dict[str, typing.Any]]
+ Proxy PUT request to provider inference endpoint
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/provider/{encode_path_param(name)}/-/{encode_path_param(trailing_uri)}",
+ method="PUT",
+ json=request,
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def provider_proxy_delete(
+ self, workspace: str, name: str, trailing_uri: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[typing.Any]:
+ """
+ Proxy requests to provider inference endpoints.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Any]
+ Proxy DELETE request to provider inference endpoint
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/provider/{encode_path_param(name)}/-/{encode_path_param(trailing_uri)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if _response is None or not _response.text.strip():
+ return AsyncHttpResponse(response=_response, data=None)
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def provider_proxy_patch(
+ self,
+ workspace: str,
+ name: str,
+ trailing_uri: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Proxy requests to provider inference endpoints.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ trailing_uri : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Dict[str, typing.Any]]
+ Proxy PATCH request to provider inference endpoint
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/provider/{encode_path_param(name)}/-/{encode_path_param(trailing_uri)}",
+ method="PATCH",
+ json=request,
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def provider_ready(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Check if a model provider is registered in the gateway's cache.
+
+ This is a lightweight endpoint that only checks the gateway's internal state,
+ without making any requests to the actual provider backend. Use this to verify
+ the gateway is ready to route requests to a provider after deployment.
+
+ Returns:
+ 200 OK with provider info if the provider is registered
+ 404 Not Found if the provider is not yet in the gateway's cache
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Dict[str, typing.Any]]
+ Check if the gateway can route to a provider
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/inference-gateway/v2/workspaces/{encode_path_param(workspace)}/provider/{encode_path_param(name)}/ready",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
diff --git a/sdks/python/ingest/__init__.py b/sdks/python/ingest/__init__.py
new file mode 100644
index 0000000000..e653e76a4a
--- /dev/null
+++ b/sdks/python/ingest/__init__.py
@@ -0,0 +1,34 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
+import typing
+from importlib import import_module
+
+if typing.TYPE_CHECKING:
+ from .types import AtifIngestRequestSchemaVersion
+_dynamic_imports: typing.Dict[str, str] = {"AtifIngestRequestSchemaVersion": ".types"}
+
+
+def __getattr__(attr_name: str) -> typing.Any:
+ module_name = _dynamic_imports.get(attr_name)
+ if module_name is None:
+ raise AttributeError(f"No {attr_name} found in _dynamic_imports for module name -> {__name__}")
+ try:
+ module = import_module(module_name, __package__)
+ if module_name == f".{attr_name}":
+ return module
+ else:
+ return getattr(module, attr_name)
+ except ImportError as e:
+ raise ImportError(f"Failed to import {attr_name} from {module_name}: {e}") from e
+ except AttributeError as e:
+ raise AttributeError(f"Failed to get {attr_name} from {module_name}: {e}") from e
+
+
+def __dir__():
+ lazy_attrs = list(_dynamic_imports.keys())
+ return sorted(lazy_attrs)
+
+
+__all__ = ["AtifIngestRequestSchemaVersion"]
diff --git a/sdks/python/ingest/client.py b/sdks/python/ingest/client.py
new file mode 100644
index 0000000000..2058439253
--- /dev/null
+++ b/sdks/python/ingest/client.py
@@ -0,0 +1,504 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.request_options import RequestOptions
+from ..types.atif_agent import AtifAgent
+from ..types.atif_final_metrics import AtifFinalMetrics
+from ..types.atif_step import AtifStep
+from ..types.captured_chat_completions_request import CapturedChatCompletionsRequest
+from ..types.captured_chat_completions_response import CapturedChatCompletionsResponse
+from ..types.chat_completions_ingest_response import ChatCompletionsIngestResponse
+from ..types.evaluation_context import EvaluationContext
+from ..types.experiment_context import ExperimentContext
+from ..types.ingest_response import IngestResponse
+from .raw_client import AsyncRawIngestClient, RawIngestClient
+from .types.atif_ingest_request_schema_version import AtifIngestRequestSchemaVersion
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class IngestClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._raw_client = RawIngestClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> RawIngestClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ RawIngestClient
+ """
+ return self._raw_client
+
+ def ingest_atif(
+ self,
+ workspace: str,
+ *,
+ schema_version: AtifIngestRequestSchemaVersion,
+ agent: AtifAgent,
+ experiment_context: typing.Optional[ExperimentContext] = OMIT,
+ evaluation_context: typing.Optional[EvaluationContext] = OMIT,
+ session_id: typing.Optional[str] = OMIT,
+ final_metrics: typing.Optional[AtifFinalMetrics] = OMIT,
+ continued_trajectory_ref: typing.Optional[str] = OMIT,
+ notes: typing.Optional[str] = OMIT,
+ extra: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ steps: typing.Optional[typing.Sequence[AtifStep]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> None:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ schema_version : AtifIngestRequestSchemaVersion
+
+ agent : AtifAgent
+
+ experiment_context : typing.Optional[ExperimentContext]
+
+ evaluation_context : typing.Optional[EvaluationContext]
+ Deprecated. Use experiment_context; when both are sent, experiment_context takes precedence.
+
+ session_id : typing.Optional[str]
+
+ final_metrics : typing.Optional[AtifFinalMetrics]
+
+ continued_trajectory_ref : typing.Optional[str]
+
+ notes : typing.Optional[str]
+
+ extra : typing.Optional[typing.Dict[str, typing.Any]]
+
+ steps : typing.Optional[typing.Sequence[AtifStep]]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ from nvidia import AtifAgent, NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.ingest.ingest_atif(
+ workspace="workspace",
+ schema_version="ATIF-v1.0",
+ agent=AtifAgent(
+ name="name",
+ version="version",
+ ),
+ )
+ """
+ _response = self._raw_client.ingest_atif(
+ workspace,
+ schema_version=schema_version,
+ agent=agent,
+ experiment_context=experiment_context,
+ evaluation_context=evaluation_context,
+ session_id=session_id,
+ final_metrics=final_metrics,
+ continued_trajectory_ref=continued_trajectory_ref,
+ notes=notes,
+ extra=extra,
+ steps=steps,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def ingest_chat_completion(
+ self,
+ workspace: str,
+ *,
+ request: CapturedChatCompletionsRequest,
+ response: CapturedChatCompletionsResponse,
+ experiment_context: typing.Optional[ExperimentContext] = OMIT,
+ evaluation_context: typing.Optional[EvaluationContext] = OMIT,
+ session_id: typing.Optional[str] = OMIT,
+ trace_id: typing.Optional[str] = OMIT,
+ provider: typing.Optional[str] = OMIT,
+ cost_usd: typing.Optional[float] = OMIT,
+ cost_input_usd: typing.Optional[float] = OMIT,
+ cost_output_usd: typing.Optional[float] = OMIT,
+ cost_details: typing.Optional[typing.Dict[str, float]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ChatCompletionsIngestResponse:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ request : CapturedChatCompletionsRequest
+
+ response : CapturedChatCompletionsResponse
+
+ experiment_context : typing.Optional[ExperimentContext]
+
+ evaluation_context : typing.Optional[EvaluationContext]
+ Deprecated. Use experiment_context; when both are sent, experiment_context takes precedence.
+
+ session_id : typing.Optional[str]
+ Groups related chat-completions calls without forcing them into the same trace.
+
+ trace_id : typing.Optional[str]
+ Opt into joining an existing trace built via OTel or ATIF. This is not a grouping mechanism for chat-completions calls; use session_id to group related calls.
+
+ provider : typing.Optional[str]
+
+ cost_usd : typing.Optional[float]
+ Total estimated cost of this model call in USD. This matches ATIF step metrics; Intake stores it as semantic cost_total_usd on spans.
+
+ cost_input_usd : typing.Optional[float]
+ Estimated input-token cost of this model call in USD.
+
+ cost_output_usd : typing.Optional[float]
+ Estimated output-token cost of this model call in USD.
+
+ cost_details : typing.Optional[typing.Dict[str, float]]
+ Additional estimated cost breakdown fields in USD.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ChatCompletionsIngestResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import (
+ CapturedChatCompletionsRequest,
+ CapturedChatMessage,
+ NvidiaApi,
+ )
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.ingest.ingest_chat_completion(
+ workspace="workspace",
+ request=CapturedChatCompletionsRequest(
+ messages=[
+ CapturedChatMessage(
+ role="user",
+ ),
+ CapturedChatMessage(
+ role="user",
+ ),
+ ],
+ model="model",
+ ),
+ response={"key": "value"},
+ )
+ """
+ _response = self._raw_client.ingest_chat_completion(
+ workspace,
+ request=request,
+ response=response,
+ experiment_context=experiment_context,
+ evaluation_context=evaluation_context,
+ session_id=session_id,
+ trace_id=trace_id,
+ provider=provider,
+ cost_usd=cost_usd,
+ cost_input_usd=cost_input_usd,
+ cost_output_usd=cost_output_usd,
+ cost_details=cost_details,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def ingest_otlp_traces(
+ self, workspace: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> IngestResponse:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ IngestResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.ingest.ingest_otlp_traces(
+ workspace="workspace",
+ )
+ """
+ _response = self._raw_client.ingest_otlp_traces(workspace, request_options=request_options)
+ return _response.data
+
+
+class AsyncIngestClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._raw_client = AsyncRawIngestClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> AsyncRawIngestClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ AsyncRawIngestClient
+ """
+ return self._raw_client
+
+ async def ingest_atif(
+ self,
+ workspace: str,
+ *,
+ schema_version: AtifIngestRequestSchemaVersion,
+ agent: AtifAgent,
+ experiment_context: typing.Optional[ExperimentContext] = OMIT,
+ evaluation_context: typing.Optional[EvaluationContext] = OMIT,
+ session_id: typing.Optional[str] = OMIT,
+ final_metrics: typing.Optional[AtifFinalMetrics] = OMIT,
+ continued_trajectory_ref: typing.Optional[str] = OMIT,
+ notes: typing.Optional[str] = OMIT,
+ extra: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ steps: typing.Optional[typing.Sequence[AtifStep]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> None:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ schema_version : AtifIngestRequestSchemaVersion
+
+ agent : AtifAgent
+
+ experiment_context : typing.Optional[ExperimentContext]
+
+ evaluation_context : typing.Optional[EvaluationContext]
+ Deprecated. Use experiment_context; when both are sent, experiment_context takes precedence.
+
+ session_id : typing.Optional[str]
+
+ final_metrics : typing.Optional[AtifFinalMetrics]
+
+ continued_trajectory_ref : typing.Optional[str]
+
+ notes : typing.Optional[str]
+
+ extra : typing.Optional[typing.Dict[str, typing.Any]]
+
+ steps : typing.Optional[typing.Sequence[AtifStep]]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi, AtifAgent
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.ingest.ingest_atif(
+ workspace="workspace",
+ schema_version="ATIF-v1.0",
+ agent=AtifAgent(
+ name="name",
+ version="version",
+ ),
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.ingest_atif(
+ workspace,
+ schema_version=schema_version,
+ agent=agent,
+ experiment_context=experiment_context,
+ evaluation_context=evaluation_context,
+ session_id=session_id,
+ final_metrics=final_metrics,
+ continued_trajectory_ref=continued_trajectory_ref,
+ notes=notes,
+ extra=extra,
+ steps=steps,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def ingest_chat_completion(
+ self,
+ workspace: str,
+ *,
+ request: CapturedChatCompletionsRequest,
+ response: CapturedChatCompletionsResponse,
+ experiment_context: typing.Optional[ExperimentContext] = OMIT,
+ evaluation_context: typing.Optional[EvaluationContext] = OMIT,
+ session_id: typing.Optional[str] = OMIT,
+ trace_id: typing.Optional[str] = OMIT,
+ provider: typing.Optional[str] = OMIT,
+ cost_usd: typing.Optional[float] = OMIT,
+ cost_input_usd: typing.Optional[float] = OMIT,
+ cost_output_usd: typing.Optional[float] = OMIT,
+ cost_details: typing.Optional[typing.Dict[str, float]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ChatCompletionsIngestResponse:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ request : CapturedChatCompletionsRequest
+
+ response : CapturedChatCompletionsResponse
+
+ experiment_context : typing.Optional[ExperimentContext]
+
+ evaluation_context : typing.Optional[EvaluationContext]
+ Deprecated. Use experiment_context; when both are sent, experiment_context takes precedence.
+
+ session_id : typing.Optional[str]
+ Groups related chat-completions calls without forcing them into the same trace.
+
+ trace_id : typing.Optional[str]
+ Opt into joining an existing trace built via OTel or ATIF. This is not a grouping mechanism for chat-completions calls; use session_id to group related calls.
+
+ provider : typing.Optional[str]
+
+ cost_usd : typing.Optional[float]
+ Total estimated cost of this model call in USD. This matches ATIF step metrics; Intake stores it as semantic cost_total_usd on spans.
+
+ cost_input_usd : typing.Optional[float]
+ Estimated input-token cost of this model call in USD.
+
+ cost_output_usd : typing.Optional[float]
+ Estimated output-token cost of this model call in USD.
+
+ cost_details : typing.Optional[typing.Dict[str, float]]
+ Additional estimated cost breakdown fields in USD.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ChatCompletionsIngestResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import (
+ AsyncNvidiaApi,
+ CapturedChatCompletionsRequest,
+ CapturedChatMessage,
+ )
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.ingest.ingest_chat_completion(
+ workspace="workspace",
+ request=CapturedChatCompletionsRequest(
+ messages=[
+ CapturedChatMessage(
+ role="user",
+ ),
+ CapturedChatMessage(
+ role="user",
+ ),
+ ],
+ model="model",
+ ),
+ response={"key": "value"},
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.ingest_chat_completion(
+ workspace,
+ request=request,
+ response=response,
+ experiment_context=experiment_context,
+ evaluation_context=evaluation_context,
+ session_id=session_id,
+ trace_id=trace_id,
+ provider=provider,
+ cost_usd=cost_usd,
+ cost_input_usd=cost_input_usd,
+ cost_output_usd=cost_output_usd,
+ cost_details=cost_details,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def ingest_otlp_traces(
+ self, workspace: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> IngestResponse:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ IngestResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.ingest.ingest_otlp_traces(
+ workspace="workspace",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.ingest_otlp_traces(workspace, request_options=request_options)
+ return _response.data
diff --git a/sdks/python/ingest/raw_client.py b/sdks/python/ingest/raw_client.py
new file mode 100644
index 0000000000..05e89aeee9
--- /dev/null
+++ b/sdks/python/ingest/raw_client.py
@@ -0,0 +1,584 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+from json.decoder import JSONDecodeError
+
+from ..core.api_error import ApiError
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.http_response import AsyncHttpResponse, HttpResponse
+from ..core.jsonable_encoder import encode_path_param
+from ..core.parse_error import ParsingError
+from ..core.pydantic_utilities import parse_obj_as
+from ..core.request_options import RequestOptions
+from ..core.serialization import convert_and_respect_annotation_metadata
+from ..errors.unprocessable_entity_error import UnprocessableEntityError
+from ..types.atif_agent import AtifAgent
+from ..types.atif_final_metrics import AtifFinalMetrics
+from ..types.atif_step import AtifStep
+from ..types.captured_chat_completions_request import CapturedChatCompletionsRequest
+from ..types.captured_chat_completions_response import CapturedChatCompletionsResponse
+from ..types.chat_completions_ingest_response import ChatCompletionsIngestResponse
+from ..types.evaluation_context import EvaluationContext
+from ..types.experiment_context import ExperimentContext
+from ..types.ingest_response import IngestResponse
+from .types.atif_ingest_request_schema_version import AtifIngestRequestSchemaVersion
+from pydantic import ValidationError
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class RawIngestClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ def ingest_atif(
+ self,
+ workspace: str,
+ *,
+ schema_version: AtifIngestRequestSchemaVersion,
+ agent: AtifAgent,
+ experiment_context: typing.Optional[ExperimentContext] = OMIT,
+ evaluation_context: typing.Optional[EvaluationContext] = OMIT,
+ session_id: typing.Optional[str] = OMIT,
+ final_metrics: typing.Optional[AtifFinalMetrics] = OMIT,
+ continued_trajectory_ref: typing.Optional[str] = OMIT,
+ notes: typing.Optional[str] = OMIT,
+ extra: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ steps: typing.Optional[typing.Sequence[AtifStep]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[None]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ schema_version : AtifIngestRequestSchemaVersion
+
+ agent : AtifAgent
+
+ experiment_context : typing.Optional[ExperimentContext]
+
+ evaluation_context : typing.Optional[EvaluationContext]
+ Deprecated. Use experiment_context; when both are sent, experiment_context takes precedence.
+
+ session_id : typing.Optional[str]
+
+ final_metrics : typing.Optional[AtifFinalMetrics]
+
+ continued_trajectory_ref : typing.Optional[str]
+
+ notes : typing.Optional[str]
+
+ extra : typing.Optional[typing.Dict[str, typing.Any]]
+
+ steps : typing.Optional[typing.Sequence[AtifStep]]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[None]
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/ingest/atif",
+ method="POST",
+ json={
+ "experiment_context": convert_and_respect_annotation_metadata(
+ object_=experiment_context, annotation=ExperimentContext, direction="write"
+ ),
+ "evaluation_context": convert_and_respect_annotation_metadata(
+ object_=evaluation_context, annotation=EvaluationContext, direction="write"
+ ),
+ "schema_version": schema_version,
+ "session_id": session_id,
+ "agent": convert_and_respect_annotation_metadata(
+ object_=agent, annotation=AtifAgent, direction="write"
+ ),
+ "final_metrics": convert_and_respect_annotation_metadata(
+ object_=final_metrics, annotation=AtifFinalMetrics, direction="write"
+ ),
+ "continued_trajectory_ref": continued_trajectory_ref,
+ "notes": notes,
+ "extra": extra,
+ "steps": convert_and_respect_annotation_metadata(
+ object_=steps, annotation=typing.Sequence[AtifStep], direction="write"
+ ),
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return HttpResponse(response=_response, data=None)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def ingest_chat_completion(
+ self,
+ workspace: str,
+ *,
+ request: CapturedChatCompletionsRequest,
+ response: CapturedChatCompletionsResponse,
+ experiment_context: typing.Optional[ExperimentContext] = OMIT,
+ evaluation_context: typing.Optional[EvaluationContext] = OMIT,
+ session_id: typing.Optional[str] = OMIT,
+ trace_id: typing.Optional[str] = OMIT,
+ provider: typing.Optional[str] = OMIT,
+ cost_usd: typing.Optional[float] = OMIT,
+ cost_input_usd: typing.Optional[float] = OMIT,
+ cost_output_usd: typing.Optional[float] = OMIT,
+ cost_details: typing.Optional[typing.Dict[str, float]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ChatCompletionsIngestResponse]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ request : CapturedChatCompletionsRequest
+
+ response : CapturedChatCompletionsResponse
+
+ experiment_context : typing.Optional[ExperimentContext]
+
+ evaluation_context : typing.Optional[EvaluationContext]
+ Deprecated. Use experiment_context; when both are sent, experiment_context takes precedence.
+
+ session_id : typing.Optional[str]
+ Groups related chat-completions calls without forcing them into the same trace.
+
+ trace_id : typing.Optional[str]
+ Opt into joining an existing trace built via OTel or ATIF. This is not a grouping mechanism for chat-completions calls; use session_id to group related calls.
+
+ provider : typing.Optional[str]
+
+ cost_usd : typing.Optional[float]
+ Total estimated cost of this model call in USD. This matches ATIF step metrics; Intake stores it as semantic cost_total_usd on spans.
+
+ cost_input_usd : typing.Optional[float]
+ Estimated input-token cost of this model call in USD.
+
+ cost_output_usd : typing.Optional[float]
+ Estimated output-token cost of this model call in USD.
+
+ cost_details : typing.Optional[typing.Dict[str, float]]
+ Additional estimated cost breakdown fields in USD.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ChatCompletionsIngestResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/ingest/chat-completions",
+ method="POST",
+ json={
+ "experiment_context": convert_and_respect_annotation_metadata(
+ object_=experiment_context, annotation=ExperimentContext, direction="write"
+ ),
+ "evaluation_context": convert_and_respect_annotation_metadata(
+ object_=evaluation_context, annotation=EvaluationContext, direction="write"
+ ),
+ "request": convert_and_respect_annotation_metadata(
+ object_=request, annotation=CapturedChatCompletionsRequest, direction="write"
+ ),
+ "response": convert_and_respect_annotation_metadata(
+ object_=response, annotation=CapturedChatCompletionsResponse, direction="write"
+ ),
+ "session_id": session_id,
+ "trace_id": trace_id,
+ "provider": provider,
+ "cost_usd": cost_usd,
+ "cost_input_usd": cost_input_usd,
+ "cost_output_usd": cost_output_usd,
+ "cost_details": cost_details,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ChatCompletionsIngestResponse,
+ parse_obj_as(
+ type_=ChatCompletionsIngestResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def ingest_otlp_traces(
+ self, workspace: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[IngestResponse]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[IngestResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/ingest/otlp/v1/traces",
+ method="POST",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ IngestResponse,
+ parse_obj_as(
+ type_=IngestResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+
+class AsyncRawIngestClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ async def ingest_atif(
+ self,
+ workspace: str,
+ *,
+ schema_version: AtifIngestRequestSchemaVersion,
+ agent: AtifAgent,
+ experiment_context: typing.Optional[ExperimentContext] = OMIT,
+ evaluation_context: typing.Optional[EvaluationContext] = OMIT,
+ session_id: typing.Optional[str] = OMIT,
+ final_metrics: typing.Optional[AtifFinalMetrics] = OMIT,
+ continued_trajectory_ref: typing.Optional[str] = OMIT,
+ notes: typing.Optional[str] = OMIT,
+ extra: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ steps: typing.Optional[typing.Sequence[AtifStep]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[None]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ schema_version : AtifIngestRequestSchemaVersion
+
+ agent : AtifAgent
+
+ experiment_context : typing.Optional[ExperimentContext]
+
+ evaluation_context : typing.Optional[EvaluationContext]
+ Deprecated. Use experiment_context; when both are sent, experiment_context takes precedence.
+
+ session_id : typing.Optional[str]
+
+ final_metrics : typing.Optional[AtifFinalMetrics]
+
+ continued_trajectory_ref : typing.Optional[str]
+
+ notes : typing.Optional[str]
+
+ extra : typing.Optional[typing.Dict[str, typing.Any]]
+
+ steps : typing.Optional[typing.Sequence[AtifStep]]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[None]
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/ingest/atif",
+ method="POST",
+ json={
+ "experiment_context": convert_and_respect_annotation_metadata(
+ object_=experiment_context, annotation=ExperimentContext, direction="write"
+ ),
+ "evaluation_context": convert_and_respect_annotation_metadata(
+ object_=evaluation_context, annotation=EvaluationContext, direction="write"
+ ),
+ "schema_version": schema_version,
+ "session_id": session_id,
+ "agent": convert_and_respect_annotation_metadata(
+ object_=agent, annotation=AtifAgent, direction="write"
+ ),
+ "final_metrics": convert_and_respect_annotation_metadata(
+ object_=final_metrics, annotation=AtifFinalMetrics, direction="write"
+ ),
+ "continued_trajectory_ref": continued_trajectory_ref,
+ "notes": notes,
+ "extra": extra,
+ "steps": convert_and_respect_annotation_metadata(
+ object_=steps, annotation=typing.Sequence[AtifStep], direction="write"
+ ),
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return AsyncHttpResponse(response=_response, data=None)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def ingest_chat_completion(
+ self,
+ workspace: str,
+ *,
+ request: CapturedChatCompletionsRequest,
+ response: CapturedChatCompletionsResponse,
+ experiment_context: typing.Optional[ExperimentContext] = OMIT,
+ evaluation_context: typing.Optional[EvaluationContext] = OMIT,
+ session_id: typing.Optional[str] = OMIT,
+ trace_id: typing.Optional[str] = OMIT,
+ provider: typing.Optional[str] = OMIT,
+ cost_usd: typing.Optional[float] = OMIT,
+ cost_input_usd: typing.Optional[float] = OMIT,
+ cost_output_usd: typing.Optional[float] = OMIT,
+ cost_details: typing.Optional[typing.Dict[str, float]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ChatCompletionsIngestResponse]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ request : CapturedChatCompletionsRequest
+
+ response : CapturedChatCompletionsResponse
+
+ experiment_context : typing.Optional[ExperimentContext]
+
+ evaluation_context : typing.Optional[EvaluationContext]
+ Deprecated. Use experiment_context; when both are sent, experiment_context takes precedence.
+
+ session_id : typing.Optional[str]
+ Groups related chat-completions calls without forcing them into the same trace.
+
+ trace_id : typing.Optional[str]
+ Opt into joining an existing trace built via OTel or ATIF. This is not a grouping mechanism for chat-completions calls; use session_id to group related calls.
+
+ provider : typing.Optional[str]
+
+ cost_usd : typing.Optional[float]
+ Total estimated cost of this model call in USD. This matches ATIF step metrics; Intake stores it as semantic cost_total_usd on spans.
+
+ cost_input_usd : typing.Optional[float]
+ Estimated input-token cost of this model call in USD.
+
+ cost_output_usd : typing.Optional[float]
+ Estimated output-token cost of this model call in USD.
+
+ cost_details : typing.Optional[typing.Dict[str, float]]
+ Additional estimated cost breakdown fields in USD.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ChatCompletionsIngestResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/ingest/chat-completions",
+ method="POST",
+ json={
+ "experiment_context": convert_and_respect_annotation_metadata(
+ object_=experiment_context, annotation=ExperimentContext, direction="write"
+ ),
+ "evaluation_context": convert_and_respect_annotation_metadata(
+ object_=evaluation_context, annotation=EvaluationContext, direction="write"
+ ),
+ "request": convert_and_respect_annotation_metadata(
+ object_=request, annotation=CapturedChatCompletionsRequest, direction="write"
+ ),
+ "response": convert_and_respect_annotation_metadata(
+ object_=response, annotation=CapturedChatCompletionsResponse, direction="write"
+ ),
+ "session_id": session_id,
+ "trace_id": trace_id,
+ "provider": provider,
+ "cost_usd": cost_usd,
+ "cost_input_usd": cost_input_usd,
+ "cost_output_usd": cost_output_usd,
+ "cost_details": cost_details,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ChatCompletionsIngestResponse,
+ parse_obj_as(
+ type_=ChatCompletionsIngestResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def ingest_otlp_traces(
+ self, workspace: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[IngestResponse]:
+ """
+ Parameters
+ ----------
+ workspace : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[IngestResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/intake/v2/workspaces/{encode_path_param(workspace)}/ingest/otlp/v1/traces",
+ method="POST",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ IngestResponse,
+ parse_obj_as(
+ type_=IngestResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
diff --git a/sdks/python/ingest/types/__init__.py b/sdks/python/ingest/types/__init__.py
new file mode 100644
index 0000000000..71eb509f44
--- /dev/null
+++ b/sdks/python/ingest/types/__init__.py
@@ -0,0 +1,34 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
+import typing
+from importlib import import_module
+
+if typing.TYPE_CHECKING:
+ from .atif_ingest_request_schema_version import AtifIngestRequestSchemaVersion
+_dynamic_imports: typing.Dict[str, str] = {"AtifIngestRequestSchemaVersion": ".atif_ingest_request_schema_version"}
+
+
+def __getattr__(attr_name: str) -> typing.Any:
+ module_name = _dynamic_imports.get(attr_name)
+ if module_name is None:
+ raise AttributeError(f"No {attr_name} found in _dynamic_imports for module name -> {__name__}")
+ try:
+ module = import_module(module_name, __package__)
+ if module_name == f".{attr_name}":
+ return module
+ else:
+ return getattr(module, attr_name)
+ except ImportError as e:
+ raise ImportError(f"Failed to import {attr_name} from {module_name}: {e}") from e
+ except AttributeError as e:
+ raise AttributeError(f"Failed to get {attr_name} from {module_name}: {e}") from e
+
+
+def __dir__():
+ lazy_attrs = list(_dynamic_imports.keys())
+ return sorted(lazy_attrs)
+
+
+__all__ = ["AtifIngestRequestSchemaVersion"]
diff --git a/sdks/python/ingest/types/atif_ingest_request_schema_version.py b/sdks/python/ingest/types/atif_ingest_request_schema_version.py
new file mode 100644
index 0000000000..3bfa4dbbea
--- /dev/null
+++ b/sdks/python/ingest/types/atif_ingest_request_schema_version.py
@@ -0,0 +1,10 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+AtifIngestRequestSchemaVersion = typing.Union[
+ typing.Literal[
+ "ATIF-v1.0", "ATIF-v1.1", "ATIF-v1.2", "ATIF-v1.3", "ATIF-v1.4", "ATIF-v1.5", "ATIF-v1.6", "ATIF-v1.7"
+ ],
+ typing.Any,
+]
diff --git a/sdks/python/jobs/__init__.py b/sdks/python/jobs/__init__.py
new file mode 100644
index 0000000000..df4abf7e8c
--- /dev/null
+++ b/sdks/python/jobs/__init__.py
@@ -0,0 +1,55 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
+import typing
+from importlib import import_module
+
+if typing.TYPE_CHECKING:
+ from .types import (
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem,
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Docker,
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_E2E,
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_KubernetesJob,
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Subprocess,
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_VolcanoJob,
+ )
+_dynamic_imports: typing.Dict[str, str] = {
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem": ".types",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Docker": ".types",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_E2E": ".types",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_KubernetesJob": ".types",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Subprocess": ".types",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_VolcanoJob": ".types",
+}
+
+
+def __getattr__(attr_name: str) -> typing.Any:
+ module_name = _dynamic_imports.get(attr_name)
+ if module_name is None:
+ raise AttributeError(f"No {attr_name} found in _dynamic_imports for module name -> {__name__}")
+ try:
+ module = import_module(module_name, __package__)
+ if module_name == f".{attr_name}":
+ return module
+ else:
+ return getattr(module, attr_name)
+ except ImportError as e:
+ raise ImportError(f"Failed to import {attr_name} from {module_name}: {e}") from e
+ except AttributeError as e:
+ raise AttributeError(f"Failed to get {attr_name} from {module_name}: {e}") from e
+
+
+def __dir__():
+ lazy_attrs = list(_dynamic_imports.keys())
+ return sorted(lazy_attrs)
+
+
+__all__ = [
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Docker",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_E2E",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_KubernetesJob",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Subprocess",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_VolcanoJob",
+]
diff --git a/sdks/python/jobs/client.py b/sdks/python/jobs/client.py
new file mode 100644
index 0000000000..da31d91cb4
--- /dev/null
+++ b/sdks/python/jobs/client.py
@@ -0,0 +1,2184 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.request_options import RequestOptions
+from ..types.file_storage_type import FileStorageType
+from ..types.platform_job_list_result_response import PlatformJobListResultResponse
+from ..types.platform_job_list_task_response import PlatformJobListTaskResponse
+from ..types.platform_job_log_page import PlatformJobLogPage
+from ..types.platform_job_response import PlatformJobResponse
+from ..types.platform_job_responses_page import PlatformJobResponsesPage
+from ..types.platform_job_result_response import PlatformJobResultResponse
+from ..types.platform_job_sort_field import PlatformJobSortField
+from ..types.platform_job_spec_input import PlatformJobSpecInput
+from ..types.platform_job_status import PlatformJobStatus
+from ..types.platform_job_status_response import PlatformJobStatusResponse
+from ..types.platform_job_step import PlatformJobStep
+from ..types.platform_job_step_with_contexts_page import PlatformJobStepWithContextsPage
+from ..types.platform_job_steps_list_filter import PlatformJobStepsListFilter
+from ..types.platform_job_task import PlatformJobTask
+from ..types.platform_jobs_list_filter import PlatformJobsListFilter
+from .raw_client import AsyncRawJobsClient, RawJobsClient
+from .types.get_execution_profiles_apis_jobs_v2execution_profiles_get_response_item import (
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem,
+)
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class JobsClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._raw_client = RawJobsClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> RawJobsClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ RawJobsClient
+ """
+ return self._raw_client
+
+ def get_execution_profiles(
+ self, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.List[GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem]:
+ """
+ Get all currently configured execution profiles.
+
+ Parameters
+ ----------
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.List[GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem]
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.get_execution_profiles()
+ """
+ _response = self._raw_client.get_execution_profiles(request_options=request_options)
+ return _response.data
+
+ def list_jobs(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[PlatformJobSortField] = None,
+ filter: typing.Optional[PlatformJobsListFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> PlatformJobResponsesPage:
+ """
+ List platform jobs with filtering and pagination.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[PlatformJobSortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[PlatformJobsListFilter]
+ Filter jobs by workspace, project, name, status, source, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobResponsesPage
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.list_jobs(
+ workspace="workspace",
+ )
+ """
+ _response = self._raw_client.list_jobs(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ def create_job(
+ self,
+ workspace: str,
+ *,
+ spec: typing.Dict[str, typing.Any],
+ platform_spec: PlatformJobSpecInput,
+ source: str,
+ name: typing.Optional[str] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ project: typing.Optional[str] = OMIT,
+ ownership: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ custom_fields: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> PlatformJobResponse:
+ """
+ Create a new platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ spec : typing.Dict[str, typing.Any]
+
+ platform_spec : PlatformJobSpecInput
+
+ source : str
+
+ name : typing.Optional[str]
+
+ description : typing.Optional[str]
+
+ project : typing.Optional[str]
+
+ ownership : typing.Optional[typing.Dict[str, typing.Any]]
+
+ custom_fields : typing.Optional[typing.Dict[str, typing.Any]]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import (
+ ContainerSpec,
+ NvidiaApi,
+ PlatformJobSpecInput,
+ PlatformJobStepSpecInput,
+ PlatformJobStepSpecInputExecutor_Cpu,
+ )
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.create_job(
+ workspace="workspace",
+ spec={"key": "value"},
+ platform_spec=PlatformJobSpecInput(
+ steps=[
+ PlatformJobStepSpecInput(
+ name="preprocess",
+ executor=PlatformJobStepSpecInputExecutor_Cpu(
+ container=ContainerSpec(
+ image="image",
+ ),
+ ),
+ )
+ ],
+ ),
+ source="source",
+ )
+ """
+ _response = self._raw_client.create_job(
+ workspace,
+ spec=spec,
+ platform_spec=platform_spec,
+ source=source,
+ name=name,
+ description=description,
+ project=project,
+ ownership=ownership,
+ custom_fields=custom_fields,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def get_job_result(
+ self, workspace: str, job: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> PlatformJobResultResponse:
+ """
+ Get a specific job result.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobResultResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.get_job_result(
+ workspace="workspace",
+ job="job",
+ name="name",
+ )
+ """
+ _response = self._raw_client.get_job_result(workspace, job, name, request_options=request_options)
+ return _response.data
+
+ def create_job_result(
+ self,
+ workspace: str,
+ job: str,
+ name: str,
+ *,
+ artifact_url: str,
+ artifact_storage_type: FileStorageType,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> PlatformJobResultResponse:
+ """
+ Create a new result for a job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ artifact_url : str
+
+ artifact_storage_type : FileStorageType
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobResultResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.create_job_result(
+ workspace="workspace",
+ job="job",
+ name="name",
+ artifact_url="artifact_url",
+ artifact_storage_type="fileset",
+ )
+ """
+ _response = self._raw_client.create_job_result(
+ workspace,
+ job,
+ name,
+ artifact_url=artifact_url,
+ artifact_storage_type=artifact_storage_type,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def download_job_result(
+ self, workspace: str, job: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Iterator[bytes]:
+ """
+ Download a job result file.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration. You can pass in configuration such as `chunk_size`, and more to customize the request and response.
+
+ Returns
+ -------
+ typing.Iterator[bytes]
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.download_job_result(
+ workspace="workspace",
+ job="job",
+ name="name",
+ )
+ """
+ with self._raw_client.download_job_result(workspace, job, name, request_options=request_options) as r:
+ yield from r.data
+
+ def get_job_step(
+ self, workspace: str, job: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> PlatformJobStep:
+ """
+ Get a specific job step.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobStep
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.get_job_step(
+ workspace="workspace",
+ job="job",
+ name="name",
+ )
+ """
+ _response = self._raw_client.get_job_step(workspace, job, name, request_options=request_options)
+ return _response.data
+
+ def update_job_step_status(
+ self,
+ workspace: str,
+ job: str,
+ name: str,
+ *,
+ status: PlatformJobStatus,
+ status_details: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ error_details: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> PlatformJobStep:
+ """
+ Update a job step status.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ status : PlatformJobStatus
+ The new status to set for the job.
+
+ status_details : typing.Optional[typing.Dict[str, typing.Any]]
+ Optional status details related to the status update.
+
+ error_details : typing.Optional[typing.Dict[str, typing.Any]]
+ Optional error details related to the status update.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobStep
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.update_job_step_status(
+ workspace="workspace",
+ job="job",
+ name="name",
+ status="created",
+ )
+ """
+ _response = self._raw_client.update_job_step_status(
+ workspace,
+ job,
+ name,
+ status=status,
+ status_details=status_details,
+ error_details=error_details,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def list_job_step_tasks(
+ self, workspace: str, job: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> PlatformJobListTaskResponse:
+ """
+ List tasks for a job step.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobListTaskResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.list_job_step_tasks(
+ workspace="workspace",
+ job="job",
+ name="name",
+ )
+ """
+ _response = self._raw_client.list_job_step_tasks(workspace, job, name, request_options=request_options)
+ return _response.data
+
+ def get_job_step_task(
+ self, workspace: str, job: str, step: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> PlatformJobTask:
+ """
+ Get a specific job step task.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ step : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobTask
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.get_job_step_task(
+ workspace="workspace",
+ job="job",
+ step="step",
+ name="name",
+ )
+ """
+ _response = self._raw_client.get_job_step_task(workspace, job, step, name, request_options=request_options)
+ return _response.data
+
+ def update_job_step_task(
+ self,
+ workspace: str,
+ job: str,
+ step: str,
+ name: str,
+ *,
+ status: typing.Optional[PlatformJobStatus] = OMIT,
+ status_details: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ error_details: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ error_stack: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> PlatformJobTask:
+ """
+ Update a job step task.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ step : str
+
+ name : str
+
+ status : typing.Optional[PlatformJobStatus]
+
+ status_details : typing.Optional[typing.Dict[str, typing.Any]]
+
+ error_details : typing.Optional[typing.Dict[str, typing.Any]]
+
+ error_stack : typing.Optional[str]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobTask
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.update_job_step_task(
+ workspace="workspace",
+ job="job",
+ step="step",
+ name="name",
+ )
+ """
+ _response = self._raw_client.update_job_step_task(
+ workspace,
+ job,
+ step,
+ name,
+ status=status,
+ status_details=status_details,
+ error_details=error_details,
+ error_stack=error_stack,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def get_job(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> PlatformJobResponse:
+ """
+ Get a platform job by name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.get_job(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.get_job(workspace, name, request_options=request_options)
+ return _response.data
+
+ def delete_job(self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None) -> None:
+ """
+ Delete a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.delete_job(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.delete_job(workspace, name, request_options=request_options)
+ return _response.data
+
+ def cancel_job(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> PlatformJobResponse:
+ """
+ Cancel a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.cancel_job(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.cancel_job(workspace, name, request_options=request_options)
+ return _response.data
+
+ def page_job_logs(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ limit: typing.Optional[int] = None,
+ page_cursor: typing.Optional[str] = None,
+ attempt_id: typing.Optional[int] = None,
+ step_id: typing.Optional[str] = None,
+ task_id: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> PlatformJobLogPage:
+ """
+ Get paginated logs for a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ limit : typing.Optional[int]
+ Maximum number of logs to return
+
+ page_cursor : typing.Optional[str]
+ Page cursor
+
+ attempt_id : typing.Optional[int]
+ Filter logs by job attempt ID
+
+ step_id : typing.Optional[str]
+ Filter logs by step name
+
+ task_id : typing.Optional[str]
+ Filter logs by task ID
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobLogPage
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.page_job_logs(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.page_job_logs(
+ workspace,
+ name,
+ limit=limit,
+ page_cursor=page_cursor,
+ attempt_id=attempt_id,
+ step_id=step_id,
+ task_id=task_id,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def pause_job(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> PlatformJobResponse:
+ """
+ Pause a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.pause_job(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.pause_job(workspace, name, request_options=request_options)
+ return _response.data
+
+ def list_job_results(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ sort: typing.Optional[PlatformJobSortField] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> PlatformJobListResultResponse:
+ """
+ List results for a job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ sort : typing.Optional[PlatformJobSortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobListResultResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.list_job_results(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.list_job_results(workspace, name, sort=sort, request_options=request_options)
+ return _response.data
+
+ def resume_job(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> PlatformJobResponse:
+ """
+ Resume a paused platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.resume_job(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.resume_job(workspace, name, request_options=request_options)
+ return _response.data
+
+ def get_job_status(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> PlatformJobStatusResponse:
+ """
+ Get the status of a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobStatusResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.get_job_status(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.get_job_status(workspace, name, request_options=request_options)
+ return _response.data
+
+ def update_job_status_details(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> typing.Any:
+ """
+ Update the status details of a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Any
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.update_job_status_details(
+ workspace="workspace",
+ name="name",
+ request={"key": "value"},
+ )
+ """
+ _response = self._raw_client.update_job_status_details(
+ workspace, name, request=request, request_options=request_options
+ )
+ return _response.data
+
+ def list_steps(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[PlatformJobSortField] = None,
+ filter: typing.Optional[PlatformJobStepsListFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> PlatformJobStepWithContextsPage:
+ """
+ List job steps with pagination and filtering.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[PlatformJobSortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[PlatformJobStepsListFilter]
+ Filter steps by job, status, and source.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobStepWithContextsPage
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.jobs.list_steps(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.list_steps(
+ workspace, name, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+
+class AsyncJobsClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._raw_client = AsyncRawJobsClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> AsyncRawJobsClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ AsyncRawJobsClient
+ """
+ return self._raw_client
+
+ async def get_execution_profiles(
+ self, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.List[GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem]:
+ """
+ Get all currently configured execution profiles.
+
+ Parameters
+ ----------
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.List[GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem]
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.get_execution_profiles()
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_execution_profiles(request_options=request_options)
+ return _response.data
+
+ async def list_jobs(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[PlatformJobSortField] = None,
+ filter: typing.Optional[PlatformJobsListFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> PlatformJobResponsesPage:
+ """
+ List platform jobs with filtering and pagination.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[PlatformJobSortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[PlatformJobsListFilter]
+ Filter jobs by workspace, project, name, status, source, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobResponsesPage
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.list_jobs(
+ workspace="workspace",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_jobs(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ async def create_job(
+ self,
+ workspace: str,
+ *,
+ spec: typing.Dict[str, typing.Any],
+ platform_spec: PlatformJobSpecInput,
+ source: str,
+ name: typing.Optional[str] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ project: typing.Optional[str] = OMIT,
+ ownership: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ custom_fields: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> PlatformJobResponse:
+ """
+ Create a new platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ spec : typing.Dict[str, typing.Any]
+
+ platform_spec : PlatformJobSpecInput
+
+ source : str
+
+ name : typing.Optional[str]
+
+ description : typing.Optional[str]
+
+ project : typing.Optional[str]
+
+ ownership : typing.Optional[typing.Dict[str, typing.Any]]
+
+ custom_fields : typing.Optional[typing.Dict[str, typing.Any]]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import (
+ AsyncNvidiaApi,
+ ContainerSpec,
+ PlatformJobSpecInput,
+ PlatformJobStepSpecInput,
+ PlatformJobStepSpecInputExecutor_Cpu,
+ )
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.create_job(
+ workspace="workspace",
+ spec={"key": "value"},
+ platform_spec=PlatformJobSpecInput(
+ steps=[
+ PlatformJobStepSpecInput(
+ name="preprocess",
+ executor=PlatformJobStepSpecInputExecutor_Cpu(
+ container=ContainerSpec(
+ image="image",
+ ),
+ ),
+ )
+ ],
+ ),
+ source="source",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.create_job(
+ workspace,
+ spec=spec,
+ platform_spec=platform_spec,
+ source=source,
+ name=name,
+ description=description,
+ project=project,
+ ownership=ownership,
+ custom_fields=custom_fields,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def get_job_result(
+ self, workspace: str, job: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> PlatformJobResultResponse:
+ """
+ Get a specific job result.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobResultResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.get_job_result(
+ workspace="workspace",
+ job="job",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_job_result(workspace, job, name, request_options=request_options)
+ return _response.data
+
+ async def create_job_result(
+ self,
+ workspace: str,
+ job: str,
+ name: str,
+ *,
+ artifact_url: str,
+ artifact_storage_type: FileStorageType,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> PlatformJobResultResponse:
+ """
+ Create a new result for a job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ artifact_url : str
+
+ artifact_storage_type : FileStorageType
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobResultResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.create_job_result(
+ workspace="workspace",
+ job="job",
+ name="name",
+ artifact_url="artifact_url",
+ artifact_storage_type="fileset",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.create_job_result(
+ workspace,
+ job,
+ name,
+ artifact_url=artifact_url,
+ artifact_storage_type=artifact_storage_type,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def download_job_result(
+ self, workspace: str, job: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.AsyncIterator[bytes]:
+ """
+ Download a job result file.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration. You can pass in configuration such as `chunk_size`, and more to customize the request and response.
+
+ Returns
+ -------
+ typing.AsyncIterator[bytes]
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.download_job_result(
+ workspace="workspace",
+ job="job",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ async with self._raw_client.download_job_result(workspace, job, name, request_options=request_options) as r:
+ async for _chunk in r.data:
+ yield _chunk
+
+ async def get_job_step(
+ self, workspace: str, job: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> PlatformJobStep:
+ """
+ Get a specific job step.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobStep
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.get_job_step(
+ workspace="workspace",
+ job="job",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_job_step(workspace, job, name, request_options=request_options)
+ return _response.data
+
+ async def update_job_step_status(
+ self,
+ workspace: str,
+ job: str,
+ name: str,
+ *,
+ status: PlatformJobStatus,
+ status_details: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ error_details: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> PlatformJobStep:
+ """
+ Update a job step status.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ status : PlatformJobStatus
+ The new status to set for the job.
+
+ status_details : typing.Optional[typing.Dict[str, typing.Any]]
+ Optional status details related to the status update.
+
+ error_details : typing.Optional[typing.Dict[str, typing.Any]]
+ Optional error details related to the status update.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobStep
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.update_job_step_status(
+ workspace="workspace",
+ job="job",
+ name="name",
+ status="created",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.update_job_step_status(
+ workspace,
+ job,
+ name,
+ status=status,
+ status_details=status_details,
+ error_details=error_details,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def list_job_step_tasks(
+ self, workspace: str, job: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> PlatformJobListTaskResponse:
+ """
+ List tasks for a job step.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobListTaskResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.list_job_step_tasks(
+ workspace="workspace",
+ job="job",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_job_step_tasks(workspace, job, name, request_options=request_options)
+ return _response.data
+
+ async def get_job_step_task(
+ self, workspace: str, job: str, step: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> PlatformJobTask:
+ """
+ Get a specific job step task.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ step : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobTask
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.get_job_step_task(
+ workspace="workspace",
+ job="job",
+ step="step",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_job_step_task(
+ workspace, job, step, name, request_options=request_options
+ )
+ return _response.data
+
+ async def update_job_step_task(
+ self,
+ workspace: str,
+ job: str,
+ step: str,
+ name: str,
+ *,
+ status: typing.Optional[PlatformJobStatus] = OMIT,
+ status_details: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ error_details: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ error_stack: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> PlatformJobTask:
+ """
+ Update a job step task.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ step : str
+
+ name : str
+
+ status : typing.Optional[PlatformJobStatus]
+
+ status_details : typing.Optional[typing.Dict[str, typing.Any]]
+
+ error_details : typing.Optional[typing.Dict[str, typing.Any]]
+
+ error_stack : typing.Optional[str]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobTask
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.update_job_step_task(
+ workspace="workspace",
+ job="job",
+ step="step",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.update_job_step_task(
+ workspace,
+ job,
+ step,
+ name,
+ status=status,
+ status_details=status_details,
+ error_details=error_details,
+ error_stack=error_stack,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def get_job(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> PlatformJobResponse:
+ """
+ Get a platform job by name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.get_job(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_job(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def delete_job(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> None:
+ """
+ Delete a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.delete_job(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.delete_job(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def cancel_job(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> PlatformJobResponse:
+ """
+ Cancel a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.cancel_job(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.cancel_job(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def page_job_logs(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ limit: typing.Optional[int] = None,
+ page_cursor: typing.Optional[str] = None,
+ attempt_id: typing.Optional[int] = None,
+ step_id: typing.Optional[str] = None,
+ task_id: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> PlatformJobLogPage:
+ """
+ Get paginated logs for a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ limit : typing.Optional[int]
+ Maximum number of logs to return
+
+ page_cursor : typing.Optional[str]
+ Page cursor
+
+ attempt_id : typing.Optional[int]
+ Filter logs by job attempt ID
+
+ step_id : typing.Optional[str]
+ Filter logs by step name
+
+ task_id : typing.Optional[str]
+ Filter logs by task ID
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobLogPage
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.page_job_logs(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.page_job_logs(
+ workspace,
+ name,
+ limit=limit,
+ page_cursor=page_cursor,
+ attempt_id=attempt_id,
+ step_id=step_id,
+ task_id=task_id,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def pause_job(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> PlatformJobResponse:
+ """
+ Pause a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.pause_job(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.pause_job(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def list_job_results(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ sort: typing.Optional[PlatformJobSortField] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> PlatformJobListResultResponse:
+ """
+ List results for a job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ sort : typing.Optional[PlatformJobSortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobListResultResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.list_job_results(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_job_results(workspace, name, sort=sort, request_options=request_options)
+ return _response.data
+
+ async def resume_job(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> PlatformJobResponse:
+ """
+ Resume a paused platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.resume_job(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.resume_job(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def get_job_status(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> PlatformJobStatusResponse:
+ """
+ Get the status of a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobStatusResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.get_job_status(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_job_status(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def update_job_status_details(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> typing.Any:
+ """
+ Update the status details of a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Any
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.update_job_status_details(
+ workspace="workspace",
+ name="name",
+ request={"key": "value"},
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.update_job_status_details(
+ workspace, name, request=request, request_options=request_options
+ )
+ return _response.data
+
+ async def list_steps(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[PlatformJobSortField] = None,
+ filter: typing.Optional[PlatformJobStepsListFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> PlatformJobStepWithContextsPage:
+ """
+ List job steps with pagination and filtering.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[PlatformJobSortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[PlatformJobStepsListFilter]
+ Filter steps by job, status, and source.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobStepWithContextsPage
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.jobs.list_steps(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_steps(
+ workspace, name, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
diff --git a/sdks/python/jobs/raw_client.py b/sdks/python/jobs/raw_client.py
new file mode 100644
index 0000000000..506d9899e7
--- /dev/null
+++ b/sdks/python/jobs/raw_client.py
@@ -0,0 +1,3211 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import contextlib
+import typing
+from json.decoder import JSONDecodeError
+
+from ..core.api_error import ApiError
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.http_response import AsyncHttpResponse, HttpResponse
+from ..core.jsonable_encoder import encode_path_param
+from ..core.parse_error import ParsingError
+from ..core.pydantic_utilities import parse_obj_as
+from ..core.request_options import RequestOptions
+from ..core.serialization import convert_and_respect_annotation_metadata
+from ..errors.conflict_error import ConflictError
+from ..errors.not_found_error import NotFoundError
+from ..errors.unprocessable_entity_error import UnprocessableEntityError
+from ..types.file_storage_type import FileStorageType
+from ..types.platform_job_list_result_response import PlatformJobListResultResponse
+from ..types.platform_job_list_task_response import PlatformJobListTaskResponse
+from ..types.platform_job_log_page import PlatformJobLogPage
+from ..types.platform_job_response import PlatformJobResponse
+from ..types.platform_job_responses_page import PlatformJobResponsesPage
+from ..types.platform_job_result_response import PlatformJobResultResponse
+from ..types.platform_job_sort_field import PlatformJobSortField
+from ..types.platform_job_spec_input import PlatformJobSpecInput
+from ..types.platform_job_status import PlatformJobStatus
+from ..types.platform_job_status_response import PlatformJobStatusResponse
+from ..types.platform_job_step import PlatformJobStep
+from ..types.platform_job_step_with_contexts_page import PlatformJobStepWithContextsPage
+from ..types.platform_job_steps_list_filter import PlatformJobStepsListFilter
+from ..types.platform_job_task import PlatformJobTask
+from ..types.platform_jobs_list_filter import PlatformJobsListFilter
+from .types.get_execution_profiles_apis_jobs_v2execution_profiles_get_response_item import (
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem,
+)
+from pydantic import ValidationError
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class RawJobsClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ def get_execution_profiles(
+ self, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[typing.List[GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem]]:
+ """
+ Get all currently configured execution profiles.
+
+ Parameters
+ ----------
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.List[GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem]]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ "apis/jobs/v2/execution-profiles",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.List[GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem],
+ parse_obj_as(
+ type_=typing.List[GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def list_jobs(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[PlatformJobSortField] = None,
+ filter: typing.Optional[PlatformJobsListFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[PlatformJobResponsesPage]:
+ """
+ List platform jobs with filtering and pagination.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[PlatformJobSortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[PlatformJobsListFilter]
+ Filter jobs by workspace, project, name, status, source, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[PlatformJobResponsesPage]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=PlatformJobsListFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobResponsesPage,
+ parse_obj_as(
+ type_=PlatformJobResponsesPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def create_job(
+ self,
+ workspace: str,
+ *,
+ spec: typing.Dict[str, typing.Any],
+ platform_spec: PlatformJobSpecInput,
+ source: str,
+ name: typing.Optional[str] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ project: typing.Optional[str] = OMIT,
+ ownership: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ custom_fields: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[PlatformJobResponse]:
+ """
+ Create a new platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ spec : typing.Dict[str, typing.Any]
+
+ platform_spec : PlatformJobSpecInput
+
+ source : str
+
+ name : typing.Optional[str]
+
+ description : typing.Optional[str]
+
+ project : typing.Optional[str]
+
+ ownership : typing.Optional[typing.Dict[str, typing.Any]]
+
+ custom_fields : typing.Optional[typing.Dict[str, typing.Any]]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[PlatformJobResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs",
+ method="POST",
+ json={
+ "name": name,
+ "description": description,
+ "project": project,
+ "spec": spec,
+ "platform_spec": convert_and_respect_annotation_metadata(
+ object_=platform_spec, annotation=PlatformJobSpecInput, direction="write"
+ ),
+ "source": source,
+ "ownership": ownership,
+ "custom_fields": custom_fields,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobResponse,
+ parse_obj_as(
+ type_=PlatformJobResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_job_result(
+ self, workspace: str, job: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[PlatformJobResultResponse]:
+ """
+ Get a specific job result.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[PlatformJobResultResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(job)}/results/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobResultResponse,
+ parse_obj_as(
+ type_=PlatformJobResultResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def create_job_result(
+ self,
+ workspace: str,
+ job: str,
+ name: str,
+ *,
+ artifact_url: str,
+ artifact_storage_type: FileStorageType,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[PlatformJobResultResponse]:
+ """
+ Create a new result for a job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ artifact_url : str
+
+ artifact_storage_type : FileStorageType
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[PlatformJobResultResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(job)}/results/{encode_path_param(name)}",
+ method="POST",
+ json={
+ "artifact_url": artifact_url,
+ "artifact_storage_type": artifact_storage_type,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobResultResponse,
+ parse_obj_as(
+ type_=PlatformJobResultResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ @contextlib.contextmanager
+ def download_job_result(
+ self, workspace: str, job: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Iterator[HttpResponse[typing.Iterator[bytes]]]:
+ """
+ Download a job result file.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration. You can pass in configuration such as `chunk_size`, and more to customize the request and response.
+
+ Returns
+ -------
+ typing.Iterator[HttpResponse[typing.Iterator[bytes]]]
+ Successful Response
+ """
+ with self._client_wrapper.httpx_client.stream(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(job)}/results/{encode_path_param(name)}/download",
+ method="GET",
+ request_options=request_options,
+ ) as _response:
+
+ def _stream() -> HttpResponse[typing.Iterator[bytes]]:
+ try:
+ if 200 <= _response.status_code < 300:
+ _chunk_size = request_options.get("chunk_size", None) if request_options is not None else None
+ return HttpResponse(
+ response=_response, data=(_chunk for _chunk in _response.iter_bytes(chunk_size=_chunk_size))
+ )
+ _response.read()
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.text
+ )
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code,
+ headers=dict(_response.headers),
+ body=_response.json(),
+ cause=e,
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ yield _stream()
+
+ def get_job_step(
+ self, workspace: str, job: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[PlatformJobStep]:
+ """
+ Get a specific job step.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[PlatformJobStep]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(job)}/steps/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobStep,
+ parse_obj_as(
+ type_=PlatformJobStep, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def update_job_step_status(
+ self,
+ workspace: str,
+ job: str,
+ name: str,
+ *,
+ status: PlatformJobStatus,
+ status_details: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ error_details: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[PlatformJobStep]:
+ """
+ Update a job step status.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ status : PlatformJobStatus
+ The new status to set for the job.
+
+ status_details : typing.Optional[typing.Dict[str, typing.Any]]
+ Optional status details related to the status update.
+
+ error_details : typing.Optional[typing.Dict[str, typing.Any]]
+ Optional error details related to the status update.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[PlatformJobStep]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(job)}/steps/{encode_path_param(name)}/status",
+ method="PATCH",
+ json={
+ "status": status,
+ "status_details": status_details,
+ "error_details": error_details,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobStep,
+ parse_obj_as(
+ type_=PlatformJobStep, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 409:
+ raise ConflictError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def list_job_step_tasks(
+ self, workspace: str, job: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[PlatformJobListTaskResponse]:
+ """
+ List tasks for a job step.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[PlatformJobListTaskResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(job)}/steps/{encode_path_param(name)}/tasks",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobListTaskResponse,
+ parse_obj_as(
+ type_=PlatformJobListTaskResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_job_step_task(
+ self, workspace: str, job: str, step: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[PlatformJobTask]:
+ """
+ Get a specific job step task.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ step : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[PlatformJobTask]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(job)}/steps/{encode_path_param(step)}/tasks/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobTask,
+ parse_obj_as(
+ type_=PlatformJobTask, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def update_job_step_task(
+ self,
+ workspace: str,
+ job: str,
+ step: str,
+ name: str,
+ *,
+ status: typing.Optional[PlatformJobStatus] = OMIT,
+ status_details: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ error_details: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ error_stack: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[PlatformJobTask]:
+ """
+ Update a job step task.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ step : str
+
+ name : str
+
+ status : typing.Optional[PlatformJobStatus]
+
+ status_details : typing.Optional[typing.Dict[str, typing.Any]]
+
+ error_details : typing.Optional[typing.Dict[str, typing.Any]]
+
+ error_stack : typing.Optional[str]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[PlatformJobTask]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(job)}/steps/{encode_path_param(step)}/tasks/{encode_path_param(name)}",
+ method="PUT",
+ json={
+ "status": status,
+ "status_details": status_details,
+ "error_details": error_details,
+ "error_stack": error_stack,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobTask,
+ parse_obj_as(
+ type_=PlatformJobTask, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_job(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[PlatformJobResponse]:
+ """
+ Get a platform job by name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[PlatformJobResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobResponse,
+ parse_obj_as(
+ type_=PlatformJobResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def delete_job(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[None]:
+ """
+ Delete a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[None]
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return HttpResponse(response=_response, data=None)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def cancel_job(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[PlatformJobResponse]:
+ """
+ Cancel a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[PlatformJobResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(name)}/cancel",
+ method="POST",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobResponse,
+ parse_obj_as(
+ type_=PlatformJobResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def page_job_logs(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ limit: typing.Optional[int] = None,
+ page_cursor: typing.Optional[str] = None,
+ attempt_id: typing.Optional[int] = None,
+ step_id: typing.Optional[str] = None,
+ task_id: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[PlatformJobLogPage]:
+ """
+ Get paginated logs for a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ limit : typing.Optional[int]
+ Maximum number of logs to return
+
+ page_cursor : typing.Optional[str]
+ Page cursor
+
+ attempt_id : typing.Optional[int]
+ Filter logs by job attempt ID
+
+ step_id : typing.Optional[str]
+ Filter logs by step name
+
+ task_id : typing.Optional[str]
+ Filter logs by task ID
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[PlatformJobLogPage]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(name)}/logs",
+ method="GET",
+ params={
+ "limit": limit,
+ "page_cursor": page_cursor,
+ "attempt_id": attempt_id,
+ "step_id": step_id,
+ "task_id": task_id,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobLogPage,
+ parse_obj_as(
+ type_=PlatformJobLogPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def pause_job(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[PlatformJobResponse]:
+ """
+ Pause a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[PlatformJobResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(name)}/pause",
+ method="POST",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobResponse,
+ parse_obj_as(
+ type_=PlatformJobResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def list_job_results(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ sort: typing.Optional[PlatformJobSortField] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[PlatformJobListResultResponse]:
+ """
+ List results for a job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ sort : typing.Optional[PlatformJobSortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[PlatformJobListResultResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(name)}/results",
+ method="GET",
+ params={
+ "sort": sort,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobListResultResponse,
+ parse_obj_as(
+ type_=PlatformJobListResultResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def resume_job(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[PlatformJobResponse]:
+ """
+ Resume a paused platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[PlatformJobResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(name)}/resume",
+ method="POST",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobResponse,
+ parse_obj_as(
+ type_=PlatformJobResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_job_status(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[PlatformJobStatusResponse]:
+ """
+ Get the status of a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[PlatformJobStatusResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(name)}/status",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobStatusResponse,
+ parse_obj_as(
+ type_=PlatformJobStatusResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def update_job_status_details(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[typing.Any]:
+ """
+ Update the status details of a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Any]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(name)}/status-details",
+ method="PATCH",
+ json=request,
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if _response is None or not _response.text.strip():
+ return HttpResponse(response=_response, data=None)
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def list_steps(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[PlatformJobSortField] = None,
+ filter: typing.Optional[PlatformJobStepsListFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[PlatformJobStepWithContextsPage]:
+ """
+ List job steps with pagination and filtering.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[PlatformJobSortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[PlatformJobStepsListFilter]
+ Filter steps by job, status, and source.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[PlatformJobStepWithContextsPage]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(name)}/steps",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=PlatformJobStepsListFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobStepWithContextsPage,
+ parse_obj_as(
+ type_=PlatformJobStepWithContextsPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+
+class AsyncRawJobsClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ async def get_execution_profiles(
+ self, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[typing.List[GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem]]:
+ """
+ Get all currently configured execution profiles.
+
+ Parameters
+ ----------
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.List[GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem]]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ "apis/jobs/v2/execution-profiles",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.List[GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem],
+ parse_obj_as(
+ type_=typing.List[GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def list_jobs(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[PlatformJobSortField] = None,
+ filter: typing.Optional[PlatformJobsListFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[PlatformJobResponsesPage]:
+ """
+ List platform jobs with filtering and pagination.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[PlatformJobSortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[PlatformJobsListFilter]
+ Filter jobs by workspace, project, name, status, source, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[PlatformJobResponsesPage]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=PlatformJobsListFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobResponsesPage,
+ parse_obj_as(
+ type_=PlatformJobResponsesPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def create_job(
+ self,
+ workspace: str,
+ *,
+ spec: typing.Dict[str, typing.Any],
+ platform_spec: PlatformJobSpecInput,
+ source: str,
+ name: typing.Optional[str] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ project: typing.Optional[str] = OMIT,
+ ownership: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ custom_fields: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[PlatformJobResponse]:
+ """
+ Create a new platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ spec : typing.Dict[str, typing.Any]
+
+ platform_spec : PlatformJobSpecInput
+
+ source : str
+
+ name : typing.Optional[str]
+
+ description : typing.Optional[str]
+
+ project : typing.Optional[str]
+
+ ownership : typing.Optional[typing.Dict[str, typing.Any]]
+
+ custom_fields : typing.Optional[typing.Dict[str, typing.Any]]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[PlatformJobResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs",
+ method="POST",
+ json={
+ "name": name,
+ "description": description,
+ "project": project,
+ "spec": spec,
+ "platform_spec": convert_and_respect_annotation_metadata(
+ object_=platform_spec, annotation=PlatformJobSpecInput, direction="write"
+ ),
+ "source": source,
+ "ownership": ownership,
+ "custom_fields": custom_fields,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobResponse,
+ parse_obj_as(
+ type_=PlatformJobResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_job_result(
+ self, workspace: str, job: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[PlatformJobResultResponse]:
+ """
+ Get a specific job result.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[PlatformJobResultResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(job)}/results/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobResultResponse,
+ parse_obj_as(
+ type_=PlatformJobResultResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def create_job_result(
+ self,
+ workspace: str,
+ job: str,
+ name: str,
+ *,
+ artifact_url: str,
+ artifact_storage_type: FileStorageType,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[PlatformJobResultResponse]:
+ """
+ Create a new result for a job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ artifact_url : str
+
+ artifact_storage_type : FileStorageType
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[PlatformJobResultResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(job)}/results/{encode_path_param(name)}",
+ method="POST",
+ json={
+ "artifact_url": artifact_url,
+ "artifact_storage_type": artifact_storage_type,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobResultResponse,
+ parse_obj_as(
+ type_=PlatformJobResultResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ @contextlib.asynccontextmanager
+ async def download_job_result(
+ self, workspace: str, job: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.AsyncIterator[AsyncHttpResponse[typing.AsyncIterator[bytes]]]:
+ """
+ Download a job result file.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration. You can pass in configuration such as `chunk_size`, and more to customize the request and response.
+
+ Returns
+ -------
+ typing.AsyncIterator[AsyncHttpResponse[typing.AsyncIterator[bytes]]]
+ Successful Response
+ """
+ async with self._client_wrapper.httpx_client.stream(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(job)}/results/{encode_path_param(name)}/download",
+ method="GET",
+ request_options=request_options,
+ ) as _response:
+
+ async def _stream() -> AsyncHttpResponse[typing.AsyncIterator[bytes]]:
+ try:
+ if 200 <= _response.status_code < 300:
+ _chunk_size = request_options.get("chunk_size", None) if request_options is not None else None
+ return AsyncHttpResponse(
+ response=_response,
+ data=(_chunk async for _chunk in _response.aiter_bytes(chunk_size=_chunk_size)),
+ )
+ await _response.aread()
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.text
+ )
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code,
+ headers=dict(_response.headers),
+ body=_response.json(),
+ cause=e,
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ yield await _stream()
+
+ async def get_job_step(
+ self, workspace: str, job: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[PlatformJobStep]:
+ """
+ Get a specific job step.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[PlatformJobStep]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(job)}/steps/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobStep,
+ parse_obj_as(
+ type_=PlatformJobStep, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def update_job_step_status(
+ self,
+ workspace: str,
+ job: str,
+ name: str,
+ *,
+ status: PlatformJobStatus,
+ status_details: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ error_details: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[PlatformJobStep]:
+ """
+ Update a job step status.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ status : PlatformJobStatus
+ The new status to set for the job.
+
+ status_details : typing.Optional[typing.Dict[str, typing.Any]]
+ Optional status details related to the status update.
+
+ error_details : typing.Optional[typing.Dict[str, typing.Any]]
+ Optional error details related to the status update.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[PlatformJobStep]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(job)}/steps/{encode_path_param(name)}/status",
+ method="PATCH",
+ json={
+ "status": status,
+ "status_details": status_details,
+ "error_details": error_details,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobStep,
+ parse_obj_as(
+ type_=PlatformJobStep, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 409:
+ raise ConflictError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def list_job_step_tasks(
+ self, workspace: str, job: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[PlatformJobListTaskResponse]:
+ """
+ List tasks for a job step.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[PlatformJobListTaskResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(job)}/steps/{encode_path_param(name)}/tasks",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobListTaskResponse,
+ parse_obj_as(
+ type_=PlatformJobListTaskResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_job_step_task(
+ self, workspace: str, job: str, step: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[PlatformJobTask]:
+ """
+ Get a specific job step task.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ step : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[PlatformJobTask]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(job)}/steps/{encode_path_param(step)}/tasks/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobTask,
+ parse_obj_as(
+ type_=PlatformJobTask, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def update_job_step_task(
+ self,
+ workspace: str,
+ job: str,
+ step: str,
+ name: str,
+ *,
+ status: typing.Optional[PlatformJobStatus] = OMIT,
+ status_details: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ error_details: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ error_stack: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[PlatformJobTask]:
+ """
+ Update a job step task.
+
+ Parameters
+ ----------
+ workspace : str
+
+ job : str
+
+ step : str
+
+ name : str
+
+ status : typing.Optional[PlatformJobStatus]
+
+ status_details : typing.Optional[typing.Dict[str, typing.Any]]
+
+ error_details : typing.Optional[typing.Dict[str, typing.Any]]
+
+ error_stack : typing.Optional[str]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[PlatformJobTask]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(job)}/steps/{encode_path_param(step)}/tasks/{encode_path_param(name)}",
+ method="PUT",
+ json={
+ "status": status,
+ "status_details": status_details,
+ "error_details": error_details,
+ "error_stack": error_stack,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobTask,
+ parse_obj_as(
+ type_=PlatformJobTask, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_job(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[PlatformJobResponse]:
+ """
+ Get a platform job by name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[PlatformJobResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobResponse,
+ parse_obj_as(
+ type_=PlatformJobResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def delete_job(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[None]:
+ """
+ Delete a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[None]
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return AsyncHttpResponse(response=_response, data=None)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def cancel_job(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[PlatformJobResponse]:
+ """
+ Cancel a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[PlatformJobResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(name)}/cancel",
+ method="POST",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobResponse,
+ parse_obj_as(
+ type_=PlatformJobResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def page_job_logs(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ limit: typing.Optional[int] = None,
+ page_cursor: typing.Optional[str] = None,
+ attempt_id: typing.Optional[int] = None,
+ step_id: typing.Optional[str] = None,
+ task_id: typing.Optional[str] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[PlatformJobLogPage]:
+ """
+ Get paginated logs for a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ limit : typing.Optional[int]
+ Maximum number of logs to return
+
+ page_cursor : typing.Optional[str]
+ Page cursor
+
+ attempt_id : typing.Optional[int]
+ Filter logs by job attempt ID
+
+ step_id : typing.Optional[str]
+ Filter logs by step name
+
+ task_id : typing.Optional[str]
+ Filter logs by task ID
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[PlatformJobLogPage]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(name)}/logs",
+ method="GET",
+ params={
+ "limit": limit,
+ "page_cursor": page_cursor,
+ "attempt_id": attempt_id,
+ "step_id": step_id,
+ "task_id": task_id,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobLogPage,
+ parse_obj_as(
+ type_=PlatformJobLogPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def pause_job(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[PlatformJobResponse]:
+ """
+ Pause a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[PlatformJobResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(name)}/pause",
+ method="POST",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobResponse,
+ parse_obj_as(
+ type_=PlatformJobResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def list_job_results(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ sort: typing.Optional[PlatformJobSortField] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[PlatformJobListResultResponse]:
+ """
+ List results for a job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ sort : typing.Optional[PlatformJobSortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[PlatformJobListResultResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(name)}/results",
+ method="GET",
+ params={
+ "sort": sort,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobListResultResponse,
+ parse_obj_as(
+ type_=PlatformJobListResultResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def resume_job(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[PlatformJobResponse]:
+ """
+ Resume a paused platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[PlatformJobResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(name)}/resume",
+ method="POST",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobResponse,
+ parse_obj_as(
+ type_=PlatformJobResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_job_status(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[PlatformJobStatusResponse]:
+ """
+ Get the status of a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[PlatformJobStatusResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(name)}/status",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobStatusResponse,
+ parse_obj_as(
+ type_=PlatformJobStatusResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def update_job_status_details(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ request: typing.Dict[str, typing.Any],
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[typing.Any]:
+ """
+ Update the status details of a platform job.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request : typing.Dict[str, typing.Any]
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Any]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(name)}/status-details",
+ method="PATCH",
+ json=request,
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if _response is None or not _response.text.strip():
+ return AsyncHttpResponse(response=_response, data=None)
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def list_steps(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[PlatformJobSortField] = None,
+ filter: typing.Optional[PlatformJobStepsListFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[PlatformJobStepWithContextsPage]:
+ """
+ List job steps with pagination and filtering.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[PlatformJobSortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[PlatformJobStepsListFilter]
+ Filter steps by job, status, and source.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[PlatformJobStepWithContextsPage]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/jobs/v2/workspaces/{encode_path_param(workspace)}/jobs/{encode_path_param(name)}/steps",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=PlatformJobStepsListFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobStepWithContextsPage,
+ parse_obj_as(
+ type_=PlatformJobStepWithContextsPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
diff --git a/sdks/python/jobs/types/__init__.py b/sdks/python/jobs/types/__init__.py
new file mode 100644
index 0000000000..98d38eb72d
--- /dev/null
+++ b/sdks/python/jobs/types/__init__.py
@@ -0,0 +1,55 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
+import typing
+from importlib import import_module
+
+if typing.TYPE_CHECKING:
+ from .get_execution_profiles_apis_jobs_v2execution_profiles_get_response_item import (
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem,
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Docker,
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_E2E,
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_KubernetesJob,
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Subprocess,
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_VolcanoJob,
+ )
+_dynamic_imports: typing.Dict[str, str] = {
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem": ".get_execution_profiles_apis_jobs_v2execution_profiles_get_response_item",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Docker": ".get_execution_profiles_apis_jobs_v2execution_profiles_get_response_item",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_E2E": ".get_execution_profiles_apis_jobs_v2execution_profiles_get_response_item",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_KubernetesJob": ".get_execution_profiles_apis_jobs_v2execution_profiles_get_response_item",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Subprocess": ".get_execution_profiles_apis_jobs_v2execution_profiles_get_response_item",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_VolcanoJob": ".get_execution_profiles_apis_jobs_v2execution_profiles_get_response_item",
+}
+
+
+def __getattr__(attr_name: str) -> typing.Any:
+ module_name = _dynamic_imports.get(attr_name)
+ if module_name is None:
+ raise AttributeError(f"No {attr_name} found in _dynamic_imports for module name -> {__name__}")
+ try:
+ module = import_module(module_name, __package__)
+ if module_name == f".{attr_name}":
+ return module
+ else:
+ return getattr(module, attr_name)
+ except ImportError as e:
+ raise ImportError(f"Failed to import {attr_name} from {module_name}: {e}") from e
+ except AttributeError as e:
+ raise AttributeError(f"Failed to get {attr_name} from {module_name}: {e}") from e
+
+
+def __dir__():
+ lazy_attrs = list(_dynamic_imports.keys())
+ return sorted(lazy_attrs)
+
+
+__all__ = [
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Docker",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_E2E",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_KubernetesJob",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Subprocess",
+ "GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_VolcanoJob",
+]
diff --git a/sdks/python/jobs/types/get_execution_profiles_apis_jobs_v2execution_profiles_get_response_item.py b/sdks/python/jobs/types/get_execution_profiles_apis_jobs_v2execution_profiles_get_response_item.py
new file mode 100644
index 0000000000..93b2e13ba5
--- /dev/null
+++ b/sdks/python/jobs/types/get_execution_profiles_apis_jobs_v2execution_profiles_get_response_item.py
@@ -0,0 +1,107 @@
+# This file was auto-generated by Fern from our API Definition.
+
+from __future__ import annotations
+
+import typing
+
+import pydantic
+import typing_extensions
+from ...core.pydantic_utilities import IS_PYDANTIC_V2, UniversalBaseModel
+from ...types.docker_job_execution_profile_config import DockerJobExecutionProfileConfig
+from ...types.job_execution_profile_config import JobExecutionProfileConfig
+from ...types.kubernetes_job_execution_profile_config import KubernetesJobExecutionProfileConfig
+from ...types.subprocess_job_execution_profile_config import SubprocessJobExecutionProfileConfig
+from ...types.subprocess_job_execution_profile_provider import SubprocessJobExecutionProfileProvider
+from ...types.volcano_job_execution_profile_config import VolcanoJobExecutionProfileConfig
+
+
+class GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Docker(UniversalBaseModel):
+ backend: typing.Literal["docker"] = "docker"
+ provider: typing.Optional[str] = None
+ profile: typing.Optional[str] = None
+ config: DockerJobExecutionProfileConfig
+
+ if IS_PYDANTIC_V2:
+ model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True) # type: ignore # Pydantic v2
+ else:
+
+ class Config:
+ frozen = True
+ smart_union = True
+ extra = pydantic.Extra.allow
+
+
+class GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_KubernetesJob(UniversalBaseModel):
+ backend: typing.Literal["kubernetes_job"] = "kubernetes_job"
+ provider: typing.Optional[str] = None
+ profile: typing.Optional[str] = None
+ config: KubernetesJobExecutionProfileConfig
+
+ if IS_PYDANTIC_V2:
+ model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True) # type: ignore # Pydantic v2
+ else:
+
+ class Config:
+ frozen = True
+ smart_union = True
+ extra = pydantic.Extra.allow
+
+
+class GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_VolcanoJob(UniversalBaseModel):
+ backend: typing.Literal["volcano_job"] = "volcano_job"
+ provider: typing.Optional[str] = None
+ profile: typing.Optional[str] = None
+ config: VolcanoJobExecutionProfileConfig
+
+ if IS_PYDANTIC_V2:
+ model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True) # type: ignore # Pydantic v2
+ else:
+
+ class Config:
+ frozen = True
+ smart_union = True
+ extra = pydantic.Extra.allow
+
+
+class GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Subprocess(UniversalBaseModel):
+ backend: typing.Literal["subprocess"] = "subprocess"
+ provider: typing.Optional[SubprocessJobExecutionProfileProvider] = None
+ profile: typing.Optional[str] = None
+ config: typing.Optional[SubprocessJobExecutionProfileConfig] = None
+
+ if IS_PYDANTIC_V2:
+ model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True) # type: ignore # Pydantic v2
+ else:
+
+ class Config:
+ frozen = True
+ smart_union = True
+ extra = pydantic.Extra.allow
+
+
+class GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_E2E(UniversalBaseModel):
+ backend: typing.Literal["e2e"] = "e2e"
+ provider: typing.Optional[str] = None
+ profile: typing.Optional[str] = None
+ config: typing.Optional[JobExecutionProfileConfig] = None
+
+ if IS_PYDANTIC_V2:
+ model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True) # type: ignore # Pydantic v2
+ else:
+
+ class Config:
+ frozen = True
+ smart_union = True
+ extra = pydantic.Extra.allow
+
+
+GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem = typing_extensions.Annotated[
+ typing.Union[
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Docker,
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_KubernetesJob,
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_VolcanoJob,
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_Subprocess,
+ GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem_E2E,
+ ],
+ pydantic.Field(discriminator="backend"),
+]
diff --git a/sdks/python/model_deployment_configs/__init__.py b/sdks/python/model_deployment_configs/__init__.py
new file mode 100644
index 0000000000..5cde0202dc
--- /dev/null
+++ b/sdks/python/model_deployment_configs/__init__.py
@@ -0,0 +1,4 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
diff --git a/sdks/python/model_deployment_configs/client.py b/sdks/python/model_deployment_configs/client.py
new file mode 100644
index 0000000000..4d1bde647d
--- /dev/null
+++ b/sdks/python/model_deployment_configs/client.py
@@ -0,0 +1,930 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.request_options import RequestOptions
+from ..types.container_executor_config import ContainerExecutorConfig
+from ..types.engine import Engine
+from ..types.model_deployment_config import ModelDeploymentConfig
+from ..types.model_deployment_config_filter import ModelDeploymentConfigFilter
+from ..types.model_deployment_config_model_spec import ModelDeploymentConfigModelSpec
+from ..types.model_deployment_configs_page import ModelDeploymentConfigsPage
+from .raw_client import AsyncRawModelDeploymentConfigsClient, RawModelDeploymentConfigsClient
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class ModelDeploymentConfigsClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._raw_client = RawModelDeploymentConfigsClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> RawModelDeploymentConfigsClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ RawModelDeploymentConfigsClient
+ """
+ return self._raw_client
+
+ def list_deployment_configs(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[str] = None,
+ filter: typing.Optional[ModelDeploymentConfigFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelDeploymentConfigsPage:
+ """
+ List ModelDeploymentConfigs for a specific workspace.
+ Returns only the latest version of each config.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[str]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[ModelDeploymentConfigFilter]
+ Filter deployment configs by workspace, project, model_entity_id, name, description, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeploymentConfigsPage
+ Return model deployment configurations for a workspace
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_deployment_configs.list_deployment_configs(
+ workspace="workspace",
+ )
+ """
+ _response = self._raw_client.list_deployment_configs(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ def create_deployment_config(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ engine: Engine,
+ model_spec: ModelDeploymentConfigModelSpec,
+ executor_config: ContainerExecutorConfig,
+ project: typing.Optional[str] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ model_entity_id: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelDeploymentConfig:
+ """
+ Create a new ModelDeploymentConfig (version 1).
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Name of the deployment configuration. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ engine : Engine
+ Inference engine selecting the compiler path (nim/vllm/generic)
+
+ model_spec : ModelDeploymentConfigModelSpec
+ What model to serve and how -- independent of the executor it runs on
+
+ executor_config : ContainerExecutorConfig
+ Compute + container settings for the executor the deployment runs on
+
+ project : typing.Optional[str]
+ The URN of the project associated with this deployment configuration
+
+ description : typing.Optional[str]
+ Optional description of the deployment configuration
+
+ model_entity_id : typing.Optional[str]
+ Optional reference to the base model entity ID for this deployment
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeploymentConfig
+ Create a new model deployment configuration
+
+ Examples
+ --------
+ from nvidia import (
+ ContainerExecutorConfig,
+ ModelDeploymentConfigModelSpec,
+ NvidiaApi,
+ )
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_deployment_configs.create_deployment_config(
+ workspace="workspace",
+ name="nim-config-v1",
+ engine="nim",
+ model_spec=ModelDeploymentConfigModelSpec(),
+ executor_config=ContainerExecutorConfig(
+ gpu=1,
+ ),
+ )
+ """
+ _response = self._raw_client.create_deployment_config(
+ workspace,
+ name=name,
+ engine=engine,
+ model_spec=model_spec,
+ executor_config=executor_config,
+ project=project,
+ description=description,
+ model_entity_id=model_entity_id,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def get_deployment_config_version(
+ self, workspace: str, config: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> ModelDeploymentConfig:
+ """
+ Get a specific version of a ModelDeploymentConfig.
+
+ Parameters
+ ----------
+ workspace : str
+
+ config : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeploymentConfig
+ Return a specific version of a model deployment configuration
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_deployment_configs.get_deployment_config_version(
+ workspace="workspace",
+ config="config",
+ name="name",
+ )
+ """
+ _response = self._raw_client.get_deployment_config_version(
+ workspace, config, name, request_options=request_options
+ )
+ return _response.data
+
+ def delete_deployment_config_version(
+ self, workspace: str, config: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> None:
+ """
+ Delete a specific version of a ModelDeploymentConfig.
+
+ This operation will fail with 409 Conflict if any ModelDeployments currently
+ reference this specific version and are not in DELETED status. Delete or wait for
+ dependent deployments to reach DELETED status before deleting the config version.
+
+ Parameters
+ ----------
+ workspace : str
+
+ config : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_deployment_configs.delete_deployment_config_version(
+ workspace="workspace",
+ config="config",
+ name="name",
+ )
+ """
+ _response = self._raw_client.delete_deployment_config_version(
+ workspace, config, name, request_options=request_options
+ )
+ return _response.data
+
+ def get_latest_deployment_config(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> ModelDeploymentConfig:
+ """
+ Get the latest version of a ModelDeploymentConfig.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeploymentConfig
+ Return the latest version of a model deployment configuration
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_deployment_configs.get_latest_deployment_config(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.get_latest_deployment_config(workspace, name, request_options=request_options)
+ return _response.data
+
+ def update_deployment_config(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ engine: Engine,
+ model_spec: ModelDeploymentConfigModelSpec,
+ executor_config: ContainerExecutorConfig,
+ description: typing.Optional[str] = OMIT,
+ model_entity_id: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelDeploymentConfig:
+ """
+ Update a ModelDeploymentConfig (creates a new immutable version).
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ engine : Engine
+ Inference engine selecting the compiler path (nim/vllm/generic)
+
+ model_spec : ModelDeploymentConfigModelSpec
+ What model to serve and how -- independent of the executor it runs on
+
+ executor_config : ContainerExecutorConfig
+ Compute + container settings for the executor the deployment runs on
+
+ description : typing.Optional[str]
+ Optional description of the deployment configuration
+
+ model_entity_id : typing.Optional[str]
+ Optional reference to the base model entity ID for this deployment
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeploymentConfig
+ Update a model deployment configuration (creates new version)
+
+ Examples
+ --------
+ from nvidia import (
+ ContainerExecutorConfig,
+ ModelDeploymentConfigModelSpec,
+ NvidiaApi,
+ )
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_deployment_configs.update_deployment_config(
+ workspace="workspace",
+ name="name",
+ engine="nim",
+ model_spec=ModelDeploymentConfigModelSpec(),
+ executor_config=ContainerExecutorConfig(
+ gpu=1,
+ ),
+ )
+ """
+ _response = self._raw_client.update_deployment_config(
+ workspace,
+ name,
+ engine=engine,
+ model_spec=model_spec,
+ executor_config=executor_config,
+ description=description,
+ model_entity_id=model_entity_id,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def delete_all_deployment_config_versions(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> None:
+ """
+ Delete all versions of a ModelDeploymentConfig.
+
+ This operation will fail with 409 Conflict if any ModelDeployments currently
+ reference this config and are not in DELETED status. Delete or wait for
+ dependent deployments to reach DELETED status before deleting the config.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_deployment_configs.delete_all_deployment_config_versions(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.delete_all_deployment_config_versions(
+ workspace, name, request_options=request_options
+ )
+ return _response.data
+
+ def list_deployment_config_versions(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.List[ModelDeploymentConfig]:
+ """
+ List all versions of a ModelDeploymentConfig.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.List[ModelDeploymentConfig]
+ Return all versions of a model deployment configuration
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_deployment_configs.list_deployment_config_versions(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.list_deployment_config_versions(workspace, name, request_options=request_options)
+ return _response.data
+
+
+class AsyncModelDeploymentConfigsClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._raw_client = AsyncRawModelDeploymentConfigsClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> AsyncRawModelDeploymentConfigsClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ AsyncRawModelDeploymentConfigsClient
+ """
+ return self._raw_client
+
+ async def list_deployment_configs(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[str] = None,
+ filter: typing.Optional[ModelDeploymentConfigFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelDeploymentConfigsPage:
+ """
+ List ModelDeploymentConfigs for a specific workspace.
+ Returns only the latest version of each config.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[str]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[ModelDeploymentConfigFilter]
+ Filter deployment configs by workspace, project, model_entity_id, name, description, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeploymentConfigsPage
+ Return model deployment configurations for a workspace
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_deployment_configs.list_deployment_configs(
+ workspace="workspace",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_deployment_configs(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ async def create_deployment_config(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ engine: Engine,
+ model_spec: ModelDeploymentConfigModelSpec,
+ executor_config: ContainerExecutorConfig,
+ project: typing.Optional[str] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ model_entity_id: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelDeploymentConfig:
+ """
+ Create a new ModelDeploymentConfig (version 1).
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Name of the deployment configuration. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ engine : Engine
+ Inference engine selecting the compiler path (nim/vllm/generic)
+
+ model_spec : ModelDeploymentConfigModelSpec
+ What model to serve and how -- independent of the executor it runs on
+
+ executor_config : ContainerExecutorConfig
+ Compute + container settings for the executor the deployment runs on
+
+ project : typing.Optional[str]
+ The URN of the project associated with this deployment configuration
+
+ description : typing.Optional[str]
+ Optional description of the deployment configuration
+
+ model_entity_id : typing.Optional[str]
+ Optional reference to the base model entity ID for this deployment
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeploymentConfig
+ Create a new model deployment configuration
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import (
+ AsyncNvidiaApi,
+ ContainerExecutorConfig,
+ ModelDeploymentConfigModelSpec,
+ )
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_deployment_configs.create_deployment_config(
+ workspace="workspace",
+ name="nim-config-v1",
+ engine="nim",
+ model_spec=ModelDeploymentConfigModelSpec(),
+ executor_config=ContainerExecutorConfig(
+ gpu=1,
+ ),
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.create_deployment_config(
+ workspace,
+ name=name,
+ engine=engine,
+ model_spec=model_spec,
+ executor_config=executor_config,
+ project=project,
+ description=description,
+ model_entity_id=model_entity_id,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def get_deployment_config_version(
+ self, workspace: str, config: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> ModelDeploymentConfig:
+ """
+ Get a specific version of a ModelDeploymentConfig.
+
+ Parameters
+ ----------
+ workspace : str
+
+ config : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeploymentConfig
+ Return a specific version of a model deployment configuration
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_deployment_configs.get_deployment_config_version(
+ workspace="workspace",
+ config="config",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_deployment_config_version(
+ workspace, config, name, request_options=request_options
+ )
+ return _response.data
+
+ async def delete_deployment_config_version(
+ self, workspace: str, config: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> None:
+ """
+ Delete a specific version of a ModelDeploymentConfig.
+
+ This operation will fail with 409 Conflict if any ModelDeployments currently
+ reference this specific version and are not in DELETED status. Delete or wait for
+ dependent deployments to reach DELETED status before deleting the config version.
+
+ Parameters
+ ----------
+ workspace : str
+
+ config : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_deployment_configs.delete_deployment_config_version(
+ workspace="workspace",
+ config="config",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.delete_deployment_config_version(
+ workspace, config, name, request_options=request_options
+ )
+ return _response.data
+
+ async def get_latest_deployment_config(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> ModelDeploymentConfig:
+ """
+ Get the latest version of a ModelDeploymentConfig.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeploymentConfig
+ Return the latest version of a model deployment configuration
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_deployment_configs.get_latest_deployment_config(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_latest_deployment_config(
+ workspace, name, request_options=request_options
+ )
+ return _response.data
+
+ async def update_deployment_config(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ engine: Engine,
+ model_spec: ModelDeploymentConfigModelSpec,
+ executor_config: ContainerExecutorConfig,
+ description: typing.Optional[str] = OMIT,
+ model_entity_id: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelDeploymentConfig:
+ """
+ Update a ModelDeploymentConfig (creates a new immutable version).
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ engine : Engine
+ Inference engine selecting the compiler path (nim/vllm/generic)
+
+ model_spec : ModelDeploymentConfigModelSpec
+ What model to serve and how -- independent of the executor it runs on
+
+ executor_config : ContainerExecutorConfig
+ Compute + container settings for the executor the deployment runs on
+
+ description : typing.Optional[str]
+ Optional description of the deployment configuration
+
+ model_entity_id : typing.Optional[str]
+ Optional reference to the base model entity ID for this deployment
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeploymentConfig
+ Update a model deployment configuration (creates new version)
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import (
+ AsyncNvidiaApi,
+ ContainerExecutorConfig,
+ ModelDeploymentConfigModelSpec,
+ )
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_deployment_configs.update_deployment_config(
+ workspace="workspace",
+ name="name",
+ engine="nim",
+ model_spec=ModelDeploymentConfigModelSpec(),
+ executor_config=ContainerExecutorConfig(
+ gpu=1,
+ ),
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.update_deployment_config(
+ workspace,
+ name,
+ engine=engine,
+ model_spec=model_spec,
+ executor_config=executor_config,
+ description=description,
+ model_entity_id=model_entity_id,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def delete_all_deployment_config_versions(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> None:
+ """
+ Delete all versions of a ModelDeploymentConfig.
+
+ This operation will fail with 409 Conflict if any ModelDeployments currently
+ reference this config and are not in DELETED status. Delete or wait for
+ dependent deployments to reach DELETED status before deleting the config.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_deployment_configs.delete_all_deployment_config_versions(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.delete_all_deployment_config_versions(
+ workspace, name, request_options=request_options
+ )
+ return _response.data
+
+ async def list_deployment_config_versions(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.List[ModelDeploymentConfig]:
+ """
+ List all versions of a ModelDeploymentConfig.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.List[ModelDeploymentConfig]
+ Return all versions of a model deployment configuration
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_deployment_configs.list_deployment_config_versions(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_deployment_config_versions(
+ workspace, name, request_options=request_options
+ )
+ return _response.data
diff --git a/sdks/python/model_deployment_configs/raw_client.py b/sdks/python/model_deployment_configs/raw_client.py
new file mode 100644
index 0000000000..9e7f3c2860
--- /dev/null
+++ b/sdks/python/model_deployment_configs/raw_client.py
@@ -0,0 +1,1218 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+from json.decoder import JSONDecodeError
+
+from ..core.api_error import ApiError
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.http_response import AsyncHttpResponse, HttpResponse
+from ..core.jsonable_encoder import encode_path_param
+from ..core.parse_error import ParsingError
+from ..core.pydantic_utilities import parse_obj_as
+from ..core.request_options import RequestOptions
+from ..core.serialization import convert_and_respect_annotation_metadata
+from ..errors.conflict_error import ConflictError
+from ..errors.not_found_error import NotFoundError
+from ..errors.unprocessable_entity_error import UnprocessableEntityError
+from ..types.container_executor_config import ContainerExecutorConfig
+from ..types.engine import Engine
+from ..types.model_deployment_config import ModelDeploymentConfig
+from ..types.model_deployment_config_filter import ModelDeploymentConfigFilter
+from ..types.model_deployment_config_model_spec import ModelDeploymentConfigModelSpec
+from ..types.model_deployment_configs_page import ModelDeploymentConfigsPage
+from pydantic import ValidationError
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class RawModelDeploymentConfigsClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ def list_deployment_configs(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[str] = None,
+ filter: typing.Optional[ModelDeploymentConfigFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ModelDeploymentConfigsPage]:
+ """
+ List ModelDeploymentConfigs for a specific workspace.
+ Returns only the latest version of each config.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[str]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[ModelDeploymentConfigFilter]
+ Filter deployment configs by workspace, project, model_entity_id, name, description, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ModelDeploymentConfigsPage]
+ Return model deployment configurations for a workspace
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployment-configs",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=ModelDeploymentConfigFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeploymentConfigsPage,
+ parse_obj_as(
+ type_=ModelDeploymentConfigsPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def create_deployment_config(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ engine: Engine,
+ model_spec: ModelDeploymentConfigModelSpec,
+ executor_config: ContainerExecutorConfig,
+ project: typing.Optional[str] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ model_entity_id: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ModelDeploymentConfig]:
+ """
+ Create a new ModelDeploymentConfig (version 1).
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Name of the deployment configuration. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ engine : Engine
+ Inference engine selecting the compiler path (nim/vllm/generic)
+
+ model_spec : ModelDeploymentConfigModelSpec
+ What model to serve and how -- independent of the executor it runs on
+
+ executor_config : ContainerExecutorConfig
+ Compute + container settings for the executor the deployment runs on
+
+ project : typing.Optional[str]
+ The URN of the project associated with this deployment configuration
+
+ description : typing.Optional[str]
+ Optional description of the deployment configuration
+
+ model_entity_id : typing.Optional[str]
+ Optional reference to the base model entity ID for this deployment
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ModelDeploymentConfig]
+ Create a new model deployment configuration
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployment-configs",
+ method="POST",
+ json={
+ "name": name,
+ "project": project,
+ "description": description,
+ "engine": engine,
+ "model_spec": convert_and_respect_annotation_metadata(
+ object_=model_spec, annotation=ModelDeploymentConfigModelSpec, direction="write"
+ ),
+ "executor_config": convert_and_respect_annotation_metadata(
+ object_=executor_config, annotation=ContainerExecutorConfig, direction="write"
+ ),
+ "model_entity_id": model_entity_id,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeploymentConfig,
+ parse_obj_as(
+ type_=ModelDeploymentConfig, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_deployment_config_version(
+ self, workspace: str, config: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[ModelDeploymentConfig]:
+ """
+ Get a specific version of a ModelDeploymentConfig.
+
+ Parameters
+ ----------
+ workspace : str
+
+ config : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ModelDeploymentConfig]
+ Return a specific version of a model deployment configuration
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployment-configs/{encode_path_param(config)}/versions/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeploymentConfig,
+ parse_obj_as(
+ type_=ModelDeploymentConfig, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def delete_deployment_config_version(
+ self, workspace: str, config: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[None]:
+ """
+ Delete a specific version of a ModelDeploymentConfig.
+
+ This operation will fail with 409 Conflict if any ModelDeployments currently
+ reference this specific version and are not in DELETED status. Delete or wait for
+ dependent deployments to reach DELETED status before deleting the config version.
+
+ Parameters
+ ----------
+ workspace : str
+
+ config : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[None]
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployment-configs/{encode_path_param(config)}/versions/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return HttpResponse(response=_response, data=None)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 409:
+ raise ConflictError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_latest_deployment_config(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[ModelDeploymentConfig]:
+ """
+ Get the latest version of a ModelDeploymentConfig.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ModelDeploymentConfig]
+ Return the latest version of a model deployment configuration
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployment-configs/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeploymentConfig,
+ parse_obj_as(
+ type_=ModelDeploymentConfig, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def update_deployment_config(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ engine: Engine,
+ model_spec: ModelDeploymentConfigModelSpec,
+ executor_config: ContainerExecutorConfig,
+ description: typing.Optional[str] = OMIT,
+ model_entity_id: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ModelDeploymentConfig]:
+ """
+ Update a ModelDeploymentConfig (creates a new immutable version).
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ engine : Engine
+ Inference engine selecting the compiler path (nim/vllm/generic)
+
+ model_spec : ModelDeploymentConfigModelSpec
+ What model to serve and how -- independent of the executor it runs on
+
+ executor_config : ContainerExecutorConfig
+ Compute + container settings for the executor the deployment runs on
+
+ description : typing.Optional[str]
+ Optional description of the deployment configuration
+
+ model_entity_id : typing.Optional[str]
+ Optional reference to the base model entity ID for this deployment
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ModelDeploymentConfig]
+ Update a model deployment configuration (creates new version)
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployment-configs/{encode_path_param(name)}",
+ method="POST",
+ json={
+ "description": description,
+ "engine": engine,
+ "model_spec": convert_and_respect_annotation_metadata(
+ object_=model_spec, annotation=ModelDeploymentConfigModelSpec, direction="write"
+ ),
+ "executor_config": convert_and_respect_annotation_metadata(
+ object_=executor_config, annotation=ContainerExecutorConfig, direction="write"
+ ),
+ "model_entity_id": model_entity_id,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeploymentConfig,
+ parse_obj_as(
+ type_=ModelDeploymentConfig, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def delete_all_deployment_config_versions(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[None]:
+ """
+ Delete all versions of a ModelDeploymentConfig.
+
+ This operation will fail with 409 Conflict if any ModelDeployments currently
+ reference this config and are not in DELETED status. Delete or wait for
+ dependent deployments to reach DELETED status before deleting the config.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[None]
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployment-configs/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return HttpResponse(response=_response, data=None)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 409:
+ raise ConflictError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def list_deployment_config_versions(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[typing.List[ModelDeploymentConfig]]:
+ """
+ List all versions of a ModelDeploymentConfig.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.List[ModelDeploymentConfig]]
+ Return all versions of a model deployment configuration
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployment-configs/{encode_path_param(name)}/versions",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.List[ModelDeploymentConfig],
+ parse_obj_as(
+ type_=typing.List[ModelDeploymentConfig], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+
+class AsyncRawModelDeploymentConfigsClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ async def list_deployment_configs(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[str] = None,
+ filter: typing.Optional[ModelDeploymentConfigFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ModelDeploymentConfigsPage]:
+ """
+ List ModelDeploymentConfigs for a specific workspace.
+ Returns only the latest version of each config.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[str]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[ModelDeploymentConfigFilter]
+ Filter deployment configs by workspace, project, model_entity_id, name, description, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ModelDeploymentConfigsPage]
+ Return model deployment configurations for a workspace
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployment-configs",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=ModelDeploymentConfigFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeploymentConfigsPage,
+ parse_obj_as(
+ type_=ModelDeploymentConfigsPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def create_deployment_config(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ engine: Engine,
+ model_spec: ModelDeploymentConfigModelSpec,
+ executor_config: ContainerExecutorConfig,
+ project: typing.Optional[str] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ model_entity_id: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ModelDeploymentConfig]:
+ """
+ Create a new ModelDeploymentConfig (version 1).
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Name of the deployment configuration. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ engine : Engine
+ Inference engine selecting the compiler path (nim/vllm/generic)
+
+ model_spec : ModelDeploymentConfigModelSpec
+ What model to serve and how -- independent of the executor it runs on
+
+ executor_config : ContainerExecutorConfig
+ Compute + container settings for the executor the deployment runs on
+
+ project : typing.Optional[str]
+ The URN of the project associated with this deployment configuration
+
+ description : typing.Optional[str]
+ Optional description of the deployment configuration
+
+ model_entity_id : typing.Optional[str]
+ Optional reference to the base model entity ID for this deployment
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ModelDeploymentConfig]
+ Create a new model deployment configuration
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployment-configs",
+ method="POST",
+ json={
+ "name": name,
+ "project": project,
+ "description": description,
+ "engine": engine,
+ "model_spec": convert_and_respect_annotation_metadata(
+ object_=model_spec, annotation=ModelDeploymentConfigModelSpec, direction="write"
+ ),
+ "executor_config": convert_and_respect_annotation_metadata(
+ object_=executor_config, annotation=ContainerExecutorConfig, direction="write"
+ ),
+ "model_entity_id": model_entity_id,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeploymentConfig,
+ parse_obj_as(
+ type_=ModelDeploymentConfig, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_deployment_config_version(
+ self, workspace: str, config: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[ModelDeploymentConfig]:
+ """
+ Get a specific version of a ModelDeploymentConfig.
+
+ Parameters
+ ----------
+ workspace : str
+
+ config : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ModelDeploymentConfig]
+ Return a specific version of a model deployment configuration
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployment-configs/{encode_path_param(config)}/versions/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeploymentConfig,
+ parse_obj_as(
+ type_=ModelDeploymentConfig, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def delete_deployment_config_version(
+ self, workspace: str, config: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[None]:
+ """
+ Delete a specific version of a ModelDeploymentConfig.
+
+ This operation will fail with 409 Conflict if any ModelDeployments currently
+ reference this specific version and are not in DELETED status. Delete or wait for
+ dependent deployments to reach DELETED status before deleting the config version.
+
+ Parameters
+ ----------
+ workspace : str
+
+ config : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[None]
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployment-configs/{encode_path_param(config)}/versions/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return AsyncHttpResponse(response=_response, data=None)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 409:
+ raise ConflictError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_latest_deployment_config(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[ModelDeploymentConfig]:
+ """
+ Get the latest version of a ModelDeploymentConfig.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ModelDeploymentConfig]
+ Return the latest version of a model deployment configuration
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployment-configs/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeploymentConfig,
+ parse_obj_as(
+ type_=ModelDeploymentConfig, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def update_deployment_config(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ engine: Engine,
+ model_spec: ModelDeploymentConfigModelSpec,
+ executor_config: ContainerExecutorConfig,
+ description: typing.Optional[str] = OMIT,
+ model_entity_id: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ModelDeploymentConfig]:
+ """
+ Update a ModelDeploymentConfig (creates a new immutable version).
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ engine : Engine
+ Inference engine selecting the compiler path (nim/vllm/generic)
+
+ model_spec : ModelDeploymentConfigModelSpec
+ What model to serve and how -- independent of the executor it runs on
+
+ executor_config : ContainerExecutorConfig
+ Compute + container settings for the executor the deployment runs on
+
+ description : typing.Optional[str]
+ Optional description of the deployment configuration
+
+ model_entity_id : typing.Optional[str]
+ Optional reference to the base model entity ID for this deployment
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ModelDeploymentConfig]
+ Update a model deployment configuration (creates new version)
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployment-configs/{encode_path_param(name)}",
+ method="POST",
+ json={
+ "description": description,
+ "engine": engine,
+ "model_spec": convert_and_respect_annotation_metadata(
+ object_=model_spec, annotation=ModelDeploymentConfigModelSpec, direction="write"
+ ),
+ "executor_config": convert_and_respect_annotation_metadata(
+ object_=executor_config, annotation=ContainerExecutorConfig, direction="write"
+ ),
+ "model_entity_id": model_entity_id,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeploymentConfig,
+ parse_obj_as(
+ type_=ModelDeploymentConfig, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def delete_all_deployment_config_versions(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[None]:
+ """
+ Delete all versions of a ModelDeploymentConfig.
+
+ This operation will fail with 409 Conflict if any ModelDeployments currently
+ reference this config and are not in DELETED status. Delete or wait for
+ dependent deployments to reach DELETED status before deleting the config.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[None]
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployment-configs/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return AsyncHttpResponse(response=_response, data=None)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 409:
+ raise ConflictError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def list_deployment_config_versions(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[typing.List[ModelDeploymentConfig]]:
+ """
+ List all versions of a ModelDeploymentConfig.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.List[ModelDeploymentConfig]]
+ Return all versions of a model deployment configuration
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployment-configs/{encode_path_param(name)}/versions",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.List[ModelDeploymentConfig],
+ parse_obj_as(
+ type_=typing.List[ModelDeploymentConfig], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
diff --git a/sdks/python/model_deployments/__init__.py b/sdks/python/model_deployments/__init__.py
new file mode 100644
index 0000000000..5cde0202dc
--- /dev/null
+++ b/sdks/python/model_deployments/__init__.py
@@ -0,0 +1,4 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
diff --git a/sdks/python/model_deployments/client.py b/sdks/python/model_deployments/client.py
new file mode 100644
index 0000000000..3a03ca3782
--- /dev/null
+++ b/sdks/python/model_deployments/client.py
@@ -0,0 +1,1106 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.request_options import RequestOptions
+from ..types.model_deployment import ModelDeployment
+from ..types.model_deployment_filter import ModelDeploymentFilter
+from ..types.model_deployment_status import ModelDeploymentStatus
+from ..types.model_deployments_page import ModelDeploymentsPage
+from .raw_client import AsyncRawModelDeploymentsClient, RawModelDeploymentsClient
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class ModelDeploymentsClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._raw_client = RawModelDeploymentsClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> RawModelDeploymentsClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ RawModelDeploymentsClient
+ """
+ return self._raw_client
+
+ def list_deployments(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[str] = None,
+ all_versions: typing.Optional[bool] = None,
+ filter: typing.Optional[ModelDeploymentFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelDeploymentsPage:
+ """
+ List ModelDeployments for a specific workspace.
+
+ By default, returns only the latest version of each deployment.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[str]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ all_versions : typing.Optional[bool]
+ If true, return all versions of each deployment. If false (default), return only the latest version.
+
+ filter : typing.Optional[ModelDeploymentFilter]
+ Filter deployments by workspace, project, status, config, model_provider_id, name, status_message, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeploymentsPage
+ Return model deployments for a workspace
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_deployments.list_deployments(
+ workspace="workspace",
+ )
+ """
+ _response = self._raw_client.list_deployments(
+ workspace,
+ page=page,
+ page_size=page_size,
+ sort=sort,
+ all_versions=all_versions,
+ filter=filter,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def create_deployment(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ config: str,
+ project: typing.Optional[str] = OMIT,
+ config_version: typing.Optional[int] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelDeployment:
+ """
+ Create a new ModelDeployment (version 1).
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Name of the deployment. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ config : str
+ Reference to the ModelDeploymentConfig name
+
+ project : typing.Optional[str]
+ The URN of the project associated with this deployment
+
+ config_version : typing.Optional[int]
+ Reference to a specific ModelDeploymentConfig version. If not specified, uses latest.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeployment
+ Create a new model deployment
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_deployments.create_deployment(
+ workspace="workspace",
+ name="llama-deploy-v1",
+ config="config",
+ )
+ """
+ _response = self._raw_client.create_deployment(
+ workspace,
+ name=name,
+ config=config,
+ project=project,
+ config_version=config_version,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def get_deployment_version(
+ self, workspace: str, deployment: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> ModelDeployment:
+ """
+ Get a specific version of a ModelDeployment.
+
+ Parameters
+ ----------
+ workspace : str
+
+ deployment : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeployment
+ Return a specific version of a model deployment
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_deployments.get_deployment_version(
+ workspace="workspace",
+ deployment="deployment",
+ name="name",
+ )
+ """
+ _response = self._raw_client.get_deployment_version(
+ workspace, deployment, name, request_options=request_options
+ )
+ return _response.data
+
+ def delete_deployment_version(
+ self, workspace: str, deployment: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Optional[typing.Any]:
+ """
+ Delete a specific version of a ModelDeployment.
+
+ If the deployment is in any state other than DELETED, this will set its status to DELETING.
+ The models controller will then:
+ 1. Delete the infrastructure (e.g., K8s NimService)
+ 2. Update the status to DELETED
+
+ If the deployment is already in DELETED status, calling delete again will permanently
+ remove it from the database.
+
+ Returns:
+ - 202 Accepted: Deployment version marked for deletion (status set to DELETING)
+ - 204 No Content: Deployment version permanently removed from database (was already DELETED)
+ - 404 Not Found: Deployment version doesn't exist
+
+ Parameters
+ ----------
+ workspace : str
+
+ deployment : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Optional[typing.Any]
+ Deployment version marked for deletion (DELETING status)
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_deployments.delete_deployment_version(
+ workspace="workspace",
+ deployment="deployment",
+ name="name",
+ )
+ """
+ _response = self._raw_client.delete_deployment_version(
+ workspace, deployment, name, request_options=request_options
+ )
+ return _response.data
+
+ def get_latest_deployment(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> ModelDeployment:
+ """
+ Get the latest version of a ModelDeployment.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeployment
+ Return the latest version of a model deployment
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_deployments.get_latest_deployment(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.get_latest_deployment(workspace, name, request_options=request_options)
+ return _response.data
+
+ def update_deployment(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ config: str,
+ config_version: typing.Optional[int] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelDeployment:
+ """
+ Update a ModelDeployment (creates a new immutable version).
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ config : str
+ Reference to the ModelDeploymentConfig name
+
+ config_version : typing.Optional[int]
+ Reference to a specific ModelDeploymentConfig version. If not specified, uses latest.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeployment
+ Update a model deployment (creates new version)
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_deployments.update_deployment(
+ workspace="workspace",
+ name="name",
+ config="config",
+ )
+ """
+ _response = self._raw_client.update_deployment(
+ workspace, name, config=config, config_version=config_version, request_options=request_options
+ )
+ return _response.data
+
+ def delete_all_deployment_versions(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Optional[typing.Any]:
+ """
+ Delete all versions of a ModelDeployment.
+
+ If the deployment is in any state other than DELETED, this will set its status to DELETING.
+ The models controller will then:
+ 1. Delete the infrastructure (e.g., K8s NimService)
+ 2. Update the status to DELETED
+
+ If the deployment is already in DELETED status, calling delete again will permanently
+ remove it from the database.
+
+ Returns:
+ - 202 Accepted: Deployment marked for deletion (status set to DELETING)
+ - 204 No Content: Deployment permanently removed from database (was already DELETED)
+ - 404 Not Found: Deployment doesn't exist
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Optional[typing.Any]
+ Deployment marked for deletion (DELETING status)
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_deployments.delete_all_deployment_versions(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.delete_all_deployment_versions(workspace, name, request_options=request_options)
+ return _response.data
+
+ def get_deployment_models(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Get Latest ModelDeployment's Model Entities from Entity Store.
+ This provides the API contract that NIMs expect from Entity Store today, for pulling LoRAs,
+ but enables us to enforce AuthZ boundaries.
+
+ TODO: Implement model entity retrieval based on deployment config.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Return model entities from Entity Store for the latest deployment
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_deployments.get_deployment_models(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.get_deployment_models(workspace, name, request_options=request_options)
+ return _response.data
+
+ def update_deployment_status(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ status: ModelDeploymentStatus,
+ version: typing.Optional[str] = None,
+ status_message: typing.Optional[str] = OMIT,
+ model_provider_id: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelDeployment:
+ """
+ Update the status of a ModelDeployment (mutable operation).
+ If version is not specified, updates the latest version.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ status : ModelDeploymentStatus
+ New status for the deployment
+
+ version : typing.Optional[str]
+
+ status_message : typing.Optional[str]
+ Detailed status message
+
+ model_provider_id : typing.Optional[str]
+ Optional reference to the auto-created ModelProvider workspace/name (format: workspace/name)
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeployment
+ Update ModelDeployment status and status_message
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_deployments.update_deployment_status(
+ workspace="workspace",
+ name="name",
+ status="UNKNOWN",
+ )
+ """
+ _response = self._raw_client.update_deployment_status(
+ workspace,
+ name,
+ status=status,
+ version=version,
+ status_message=status_message,
+ model_provider_id=model_provider_id,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def list_deployment_versions(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.List[ModelDeployment]:
+ """
+ List all versions of a ModelDeployment.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.List[ModelDeployment]
+ Return all versions of a model deployment
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_deployments.list_deployment_versions(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.list_deployment_versions(workspace, name, request_options=request_options)
+ return _response.data
+
+
+class AsyncModelDeploymentsClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._raw_client = AsyncRawModelDeploymentsClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> AsyncRawModelDeploymentsClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ AsyncRawModelDeploymentsClient
+ """
+ return self._raw_client
+
+ async def list_deployments(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[str] = None,
+ all_versions: typing.Optional[bool] = None,
+ filter: typing.Optional[ModelDeploymentFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelDeploymentsPage:
+ """
+ List ModelDeployments for a specific workspace.
+
+ By default, returns only the latest version of each deployment.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[str]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ all_versions : typing.Optional[bool]
+ If true, return all versions of each deployment. If false (default), return only the latest version.
+
+ filter : typing.Optional[ModelDeploymentFilter]
+ Filter deployments by workspace, project, status, config, model_provider_id, name, status_message, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeploymentsPage
+ Return model deployments for a workspace
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_deployments.list_deployments(
+ workspace="workspace",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_deployments(
+ workspace,
+ page=page,
+ page_size=page_size,
+ sort=sort,
+ all_versions=all_versions,
+ filter=filter,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def create_deployment(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ config: str,
+ project: typing.Optional[str] = OMIT,
+ config_version: typing.Optional[int] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelDeployment:
+ """
+ Create a new ModelDeployment (version 1).
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Name of the deployment. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ config : str
+ Reference to the ModelDeploymentConfig name
+
+ project : typing.Optional[str]
+ The URN of the project associated with this deployment
+
+ config_version : typing.Optional[int]
+ Reference to a specific ModelDeploymentConfig version. If not specified, uses latest.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeployment
+ Create a new model deployment
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_deployments.create_deployment(
+ workspace="workspace",
+ name="llama-deploy-v1",
+ config="config",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.create_deployment(
+ workspace,
+ name=name,
+ config=config,
+ project=project,
+ config_version=config_version,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def get_deployment_version(
+ self, workspace: str, deployment: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> ModelDeployment:
+ """
+ Get a specific version of a ModelDeployment.
+
+ Parameters
+ ----------
+ workspace : str
+
+ deployment : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeployment
+ Return a specific version of a model deployment
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_deployments.get_deployment_version(
+ workspace="workspace",
+ deployment="deployment",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_deployment_version(
+ workspace, deployment, name, request_options=request_options
+ )
+ return _response.data
+
+ async def delete_deployment_version(
+ self, workspace: str, deployment: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Optional[typing.Any]:
+ """
+ Delete a specific version of a ModelDeployment.
+
+ If the deployment is in any state other than DELETED, this will set its status to DELETING.
+ The models controller will then:
+ 1. Delete the infrastructure (e.g., K8s NimService)
+ 2. Update the status to DELETED
+
+ If the deployment is already in DELETED status, calling delete again will permanently
+ remove it from the database.
+
+ Returns:
+ - 202 Accepted: Deployment version marked for deletion (status set to DELETING)
+ - 204 No Content: Deployment version permanently removed from database (was already DELETED)
+ - 404 Not Found: Deployment version doesn't exist
+
+ Parameters
+ ----------
+ workspace : str
+
+ deployment : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Optional[typing.Any]
+ Deployment version marked for deletion (DELETING status)
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_deployments.delete_deployment_version(
+ workspace="workspace",
+ deployment="deployment",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.delete_deployment_version(
+ workspace, deployment, name, request_options=request_options
+ )
+ return _response.data
+
+ async def get_latest_deployment(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> ModelDeployment:
+ """
+ Get the latest version of a ModelDeployment.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeployment
+ Return the latest version of a model deployment
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_deployments.get_latest_deployment(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_latest_deployment(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def update_deployment(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ config: str,
+ config_version: typing.Optional[int] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelDeployment:
+ """
+ Update a ModelDeployment (creates a new immutable version).
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ config : str
+ Reference to the ModelDeploymentConfig name
+
+ config_version : typing.Optional[int]
+ Reference to a specific ModelDeploymentConfig version. If not specified, uses latest.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeployment
+ Update a model deployment (creates new version)
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_deployments.update_deployment(
+ workspace="workspace",
+ name="name",
+ config="config",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.update_deployment(
+ workspace, name, config=config, config_version=config_version, request_options=request_options
+ )
+ return _response.data
+
+ async def delete_all_deployment_versions(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Optional[typing.Any]:
+ """
+ Delete all versions of a ModelDeployment.
+
+ If the deployment is in any state other than DELETED, this will set its status to DELETING.
+ The models controller will then:
+ 1. Delete the infrastructure (e.g., K8s NimService)
+ 2. Update the status to DELETED
+
+ If the deployment is already in DELETED status, calling delete again will permanently
+ remove it from the database.
+
+ Returns:
+ - 202 Accepted: Deployment marked for deletion (status set to DELETING)
+ - 204 No Content: Deployment permanently removed from database (was already DELETED)
+ - 404 Not Found: Deployment doesn't exist
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Optional[typing.Any]
+ Deployment marked for deletion (DELETING status)
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_deployments.delete_all_deployment_versions(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.delete_all_deployment_versions(
+ workspace, name, request_options=request_options
+ )
+ return _response.data
+
+ async def get_deployment_models(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.Dict[str, typing.Any]:
+ """
+ Get Latest ModelDeployment's Model Entities from Entity Store.
+ This provides the API contract that NIMs expect from Entity Store today, for pulling LoRAs,
+ but enables us to enforce AuthZ boundaries.
+
+ TODO: Implement model entity retrieval based on deployment config.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.Dict[str, typing.Any]
+ Return model entities from Entity Store for the latest deployment
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_deployments.get_deployment_models(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_deployment_models(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def update_deployment_status(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ status: ModelDeploymentStatus,
+ version: typing.Optional[str] = None,
+ status_message: typing.Optional[str] = OMIT,
+ model_provider_id: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelDeployment:
+ """
+ Update the status of a ModelDeployment (mutable operation).
+ If version is not specified, updates the latest version.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ status : ModelDeploymentStatus
+ New status for the deployment
+
+ version : typing.Optional[str]
+
+ status_message : typing.Optional[str]
+ Detailed status message
+
+ model_provider_id : typing.Optional[str]
+ Optional reference to the auto-created ModelProvider workspace/name (format: workspace/name)
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelDeployment
+ Update ModelDeployment status and status_message
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_deployments.update_deployment_status(
+ workspace="workspace",
+ name="name",
+ status="UNKNOWN",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.update_deployment_status(
+ workspace,
+ name,
+ status=status,
+ version=version,
+ status_message=status_message,
+ model_provider_id=model_provider_id,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def list_deployment_versions(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> typing.List[ModelDeployment]:
+ """
+ List all versions of a ModelDeployment.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ typing.List[ModelDeployment]
+ Return all versions of a model deployment
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_deployments.list_deployment_versions(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_deployment_versions(workspace, name, request_options=request_options)
+ return _response.data
diff --git a/sdks/python/model_deployments/raw_client.py b/sdks/python/model_deployments/raw_client.py
new file mode 100644
index 0000000000..2d9f1250d7
--- /dev/null
+++ b/sdks/python/model_deployments/raw_client.py
@@ -0,0 +1,1475 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+from json.decoder import JSONDecodeError
+
+from ..core.api_error import ApiError
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.http_response import AsyncHttpResponse, HttpResponse
+from ..core.jsonable_encoder import encode_path_param
+from ..core.parse_error import ParsingError
+from ..core.pydantic_utilities import parse_obj_as
+from ..core.request_options import RequestOptions
+from ..core.serialization import convert_and_respect_annotation_metadata
+from ..errors.not_found_error import NotFoundError
+from ..errors.unprocessable_entity_error import UnprocessableEntityError
+from ..types.model_deployment import ModelDeployment
+from ..types.model_deployment_filter import ModelDeploymentFilter
+from ..types.model_deployment_status import ModelDeploymentStatus
+from ..types.model_deployments_page import ModelDeploymentsPage
+from pydantic import ValidationError
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class RawModelDeploymentsClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ def list_deployments(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[str] = None,
+ all_versions: typing.Optional[bool] = None,
+ filter: typing.Optional[ModelDeploymentFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ModelDeploymentsPage]:
+ """
+ List ModelDeployments for a specific workspace.
+
+ By default, returns only the latest version of each deployment.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[str]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ all_versions : typing.Optional[bool]
+ If true, return all versions of each deployment. If false (default), return only the latest version.
+
+ filter : typing.Optional[ModelDeploymentFilter]
+ Filter deployments by workspace, project, status, config, model_provider_id, name, status_message, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ModelDeploymentsPage]
+ Return model deployments for a workspace
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployments",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "all_versions": all_versions,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=ModelDeploymentFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeploymentsPage,
+ parse_obj_as(
+ type_=ModelDeploymentsPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def create_deployment(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ config: str,
+ project: typing.Optional[str] = OMIT,
+ config_version: typing.Optional[int] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ModelDeployment]:
+ """
+ Create a new ModelDeployment (version 1).
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Name of the deployment. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ config : str
+ Reference to the ModelDeploymentConfig name
+
+ project : typing.Optional[str]
+ The URN of the project associated with this deployment
+
+ config_version : typing.Optional[int]
+ Reference to a specific ModelDeploymentConfig version. If not specified, uses latest.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ModelDeployment]
+ Create a new model deployment
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployments",
+ method="POST",
+ json={
+ "name": name,
+ "project": project,
+ "config": config,
+ "config_version": config_version,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeployment,
+ parse_obj_as(
+ type_=ModelDeployment, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_deployment_version(
+ self, workspace: str, deployment: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[ModelDeployment]:
+ """
+ Get a specific version of a ModelDeployment.
+
+ Parameters
+ ----------
+ workspace : str
+
+ deployment : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ModelDeployment]
+ Return a specific version of a model deployment
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployments/{encode_path_param(deployment)}/versions/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeployment,
+ parse_obj_as(
+ type_=ModelDeployment, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def delete_deployment_version(
+ self, workspace: str, deployment: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[typing.Optional[typing.Any]]:
+ """
+ Delete a specific version of a ModelDeployment.
+
+ If the deployment is in any state other than DELETED, this will set its status to DELETING.
+ The models controller will then:
+ 1. Delete the infrastructure (e.g., K8s NimService)
+ 2. Update the status to DELETED
+
+ If the deployment is already in DELETED status, calling delete again will permanently
+ remove it from the database.
+
+ Returns:
+ - 202 Accepted: Deployment version marked for deletion (status set to DELETING)
+ - 204 No Content: Deployment version permanently removed from database (was already DELETED)
+ - 404 Not Found: Deployment version doesn't exist
+
+ Parameters
+ ----------
+ workspace : str
+
+ deployment : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Optional[typing.Any]]
+ Deployment version marked for deletion (DELETING status)
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployments/{encode_path_param(deployment)}/versions/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if _response is None or not _response.text.strip():
+ return HttpResponse(response=_response, data=None)
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Optional[typing.Any],
+ parse_obj_as(
+ type_=typing.Optional[typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_latest_deployment(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[ModelDeployment]:
+ """
+ Get the latest version of a ModelDeployment.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ModelDeployment]
+ Return the latest version of a model deployment
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployments/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeployment,
+ parse_obj_as(
+ type_=ModelDeployment, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def update_deployment(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ config: str,
+ config_version: typing.Optional[int] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ModelDeployment]:
+ """
+ Update a ModelDeployment (creates a new immutable version).
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ config : str
+ Reference to the ModelDeploymentConfig name
+
+ config_version : typing.Optional[int]
+ Reference to a specific ModelDeploymentConfig version. If not specified, uses latest.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ModelDeployment]
+ Update a model deployment (creates new version)
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployments/{encode_path_param(name)}",
+ method="POST",
+ json={
+ "config": config,
+ "config_version": config_version,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeployment,
+ parse_obj_as(
+ type_=ModelDeployment, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def delete_all_deployment_versions(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[typing.Optional[typing.Any]]:
+ """
+ Delete all versions of a ModelDeployment.
+
+ If the deployment is in any state other than DELETED, this will set its status to DELETING.
+ The models controller will then:
+ 1. Delete the infrastructure (e.g., K8s NimService)
+ 2. Update the status to DELETED
+
+ If the deployment is already in DELETED status, calling delete again will permanently
+ remove it from the database.
+
+ Returns:
+ - 202 Accepted: Deployment marked for deletion (status set to DELETING)
+ - 204 No Content: Deployment permanently removed from database (was already DELETED)
+ - 404 Not Found: Deployment doesn't exist
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Optional[typing.Any]]
+ Deployment marked for deletion (DELETING status)
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployments/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if _response is None or not _response.text.strip():
+ return HttpResponse(response=_response, data=None)
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Optional[typing.Any],
+ parse_obj_as(
+ type_=typing.Optional[typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_deployment_models(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Get Latest ModelDeployment's Model Entities from Entity Store.
+ This provides the API contract that NIMs expect from Entity Store today, for pulling LoRAs,
+ but enables us to enforce AuthZ boundaries.
+
+ TODO: Implement model entity retrieval based on deployment config.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.Dict[str, typing.Any]]
+ Return model entities from Entity Store for the latest deployment
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployments/{encode_path_param(name)}/models",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def update_deployment_status(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ status: ModelDeploymentStatus,
+ version: typing.Optional[str] = None,
+ status_message: typing.Optional[str] = OMIT,
+ model_provider_id: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ModelDeployment]:
+ """
+ Update the status of a ModelDeployment (mutable operation).
+ If version is not specified, updates the latest version.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ status : ModelDeploymentStatus
+ New status for the deployment
+
+ version : typing.Optional[str]
+
+ status_message : typing.Optional[str]
+ Detailed status message
+
+ model_provider_id : typing.Optional[str]
+ Optional reference to the auto-created ModelProvider workspace/name (format: workspace/name)
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ModelDeployment]
+ Update ModelDeployment status and status_message
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployments/{encode_path_param(name)}/status",
+ method="POST",
+ params={
+ "version": version,
+ },
+ json={
+ "status": status,
+ "status_message": status_message,
+ "model_provider_id": model_provider_id,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeployment,
+ parse_obj_as(
+ type_=ModelDeployment, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def list_deployment_versions(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[typing.List[ModelDeployment]]:
+ """
+ List all versions of a ModelDeployment.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[typing.List[ModelDeployment]]
+ Return all versions of a model deployment
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployments/{encode_path_param(name)}/versions",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.List[ModelDeployment],
+ parse_obj_as(
+ type_=typing.List[ModelDeployment], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+
+class AsyncRawModelDeploymentsClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ async def list_deployments(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[str] = None,
+ all_versions: typing.Optional[bool] = None,
+ filter: typing.Optional[ModelDeploymentFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ModelDeploymentsPage]:
+ """
+ List ModelDeployments for a specific workspace.
+
+ By default, returns only the latest version of each deployment.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[str]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ all_versions : typing.Optional[bool]
+ If true, return all versions of each deployment. If false (default), return only the latest version.
+
+ filter : typing.Optional[ModelDeploymentFilter]
+ Filter deployments by workspace, project, status, config, model_provider_id, name, status_message, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ModelDeploymentsPage]
+ Return model deployments for a workspace
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployments",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "all_versions": all_versions,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=ModelDeploymentFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeploymentsPage,
+ parse_obj_as(
+ type_=ModelDeploymentsPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def create_deployment(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ config: str,
+ project: typing.Optional[str] = OMIT,
+ config_version: typing.Optional[int] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ModelDeployment]:
+ """
+ Create a new ModelDeployment (version 1).
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Name of the deployment. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ config : str
+ Reference to the ModelDeploymentConfig name
+
+ project : typing.Optional[str]
+ The URN of the project associated with this deployment
+
+ config_version : typing.Optional[int]
+ Reference to a specific ModelDeploymentConfig version. If not specified, uses latest.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ModelDeployment]
+ Create a new model deployment
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployments",
+ method="POST",
+ json={
+ "name": name,
+ "project": project,
+ "config": config,
+ "config_version": config_version,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeployment,
+ parse_obj_as(
+ type_=ModelDeployment, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_deployment_version(
+ self, workspace: str, deployment: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[ModelDeployment]:
+ """
+ Get a specific version of a ModelDeployment.
+
+ Parameters
+ ----------
+ workspace : str
+
+ deployment : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ModelDeployment]
+ Return a specific version of a model deployment
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployments/{encode_path_param(deployment)}/versions/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeployment,
+ parse_obj_as(
+ type_=ModelDeployment, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def delete_deployment_version(
+ self, workspace: str, deployment: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[typing.Optional[typing.Any]]:
+ """
+ Delete a specific version of a ModelDeployment.
+
+ If the deployment is in any state other than DELETED, this will set its status to DELETING.
+ The models controller will then:
+ 1. Delete the infrastructure (e.g., K8s NimService)
+ 2. Update the status to DELETED
+
+ If the deployment is already in DELETED status, calling delete again will permanently
+ remove it from the database.
+
+ Returns:
+ - 202 Accepted: Deployment version marked for deletion (status set to DELETING)
+ - 204 No Content: Deployment version permanently removed from database (was already DELETED)
+ - 404 Not Found: Deployment version doesn't exist
+
+ Parameters
+ ----------
+ workspace : str
+
+ deployment : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Optional[typing.Any]]
+ Deployment version marked for deletion (DELETING status)
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployments/{encode_path_param(deployment)}/versions/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if _response is None or not _response.text.strip():
+ return AsyncHttpResponse(response=_response, data=None)
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Optional[typing.Any],
+ parse_obj_as(
+ type_=typing.Optional[typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_latest_deployment(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[ModelDeployment]:
+ """
+ Get the latest version of a ModelDeployment.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ModelDeployment]
+ Return the latest version of a model deployment
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployments/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeployment,
+ parse_obj_as(
+ type_=ModelDeployment, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def update_deployment(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ config: str,
+ config_version: typing.Optional[int] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ModelDeployment]:
+ """
+ Update a ModelDeployment (creates a new immutable version).
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ config : str
+ Reference to the ModelDeploymentConfig name
+
+ config_version : typing.Optional[int]
+ Reference to a specific ModelDeploymentConfig version. If not specified, uses latest.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ModelDeployment]
+ Update a model deployment (creates new version)
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployments/{encode_path_param(name)}",
+ method="POST",
+ json={
+ "config": config,
+ "config_version": config_version,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeployment,
+ parse_obj_as(
+ type_=ModelDeployment, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def delete_all_deployment_versions(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[typing.Optional[typing.Any]]:
+ """
+ Delete all versions of a ModelDeployment.
+
+ If the deployment is in any state other than DELETED, this will set its status to DELETING.
+ The models controller will then:
+ 1. Delete the infrastructure (e.g., K8s NimService)
+ 2. Update the status to DELETED
+
+ If the deployment is already in DELETED status, calling delete again will permanently
+ remove it from the database.
+
+ Returns:
+ - 202 Accepted: Deployment marked for deletion (status set to DELETING)
+ - 204 No Content: Deployment permanently removed from database (was already DELETED)
+ - 404 Not Found: Deployment doesn't exist
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Optional[typing.Any]]
+ Deployment marked for deletion (DELETING status)
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployments/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if _response is None or not _response.text.strip():
+ return AsyncHttpResponse(response=_response, data=None)
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Optional[typing.Any],
+ parse_obj_as(
+ type_=typing.Optional[typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 404:
+ raise NotFoundError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_deployment_models(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[typing.Dict[str, typing.Any]]:
+ """
+ Get Latest ModelDeployment's Model Entities from Entity Store.
+ This provides the API contract that NIMs expect from Entity Store today, for pulling LoRAs,
+ but enables us to enforce AuthZ boundaries.
+
+ TODO: Implement model entity retrieval based on deployment config.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.Dict[str, typing.Any]]
+ Return model entities from Entity Store for the latest deployment
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployments/{encode_path_param(name)}/models",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.Dict[str, typing.Any],
+ parse_obj_as(
+ type_=typing.Dict[str, typing.Any], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def update_deployment_status(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ status: ModelDeploymentStatus,
+ version: typing.Optional[str] = None,
+ status_message: typing.Optional[str] = OMIT,
+ model_provider_id: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ModelDeployment]:
+ """
+ Update the status of a ModelDeployment (mutable operation).
+ If version is not specified, updates the latest version.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ status : ModelDeploymentStatus
+ New status for the deployment
+
+ version : typing.Optional[str]
+
+ status_message : typing.Optional[str]
+ Detailed status message
+
+ model_provider_id : typing.Optional[str]
+ Optional reference to the auto-created ModelProvider workspace/name (format: workspace/name)
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ModelDeployment]
+ Update ModelDeployment status and status_message
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployments/{encode_path_param(name)}/status",
+ method="POST",
+ params={
+ "version": version,
+ },
+ json={
+ "status": status,
+ "status_message": status_message,
+ "model_provider_id": model_provider_id,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelDeployment,
+ parse_obj_as(
+ type_=ModelDeployment, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def list_deployment_versions(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[typing.List[ModelDeployment]]:
+ """
+ List all versions of a ModelDeployment.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[typing.List[ModelDeployment]]
+ Return all versions of a model deployment
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/deployments/{encode_path_param(name)}/versions",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ typing.List[ModelDeployment],
+ parse_obj_as(
+ type_=typing.List[ModelDeployment], # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
diff --git a/sdks/python/model_providers/__init__.py b/sdks/python/model_providers/__init__.py
new file mode 100644
index 0000000000..5cde0202dc
--- /dev/null
+++ b/sdks/python/model_providers/__init__.py
@@ -0,0 +1,4 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
diff --git a/sdks/python/model_providers/client.py b/sdks/python/model_providers/client.py
new file mode 100644
index 0000000000..9eb81f600e
--- /dev/null
+++ b/sdks/python/model_providers/client.py
@@ -0,0 +1,920 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.request_options import RequestOptions
+from ..types.model_provider import ModelProvider
+from ..types.model_provider_filter import ModelProviderFilter
+from ..types.model_provider_sort import ModelProviderSort
+from ..types.model_provider_status import ModelProviderStatus
+from ..types.model_providers_page import ModelProvidersPage
+from ..types.served_model_mapping import ServedModelMapping
+from .raw_client import AsyncRawModelProvidersClient, RawModelProvidersClient
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class ModelProvidersClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._raw_client = RawModelProvidersClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> RawModelProvidersClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ RawModelProvidersClient
+ """
+ return self._raw_client
+
+ def list_providers(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[ModelProviderSort] = None,
+ filter: typing.Optional[ModelProviderFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelProvidersPage:
+ """
+ List model providers for a specific workspace.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[ModelProviderSort]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[ModelProviderFilter]
+ Filter model providers by workspace, project, status, model_deployment_id, name, description, host_url, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelProvidersPage
+ Return model providers for a workspace
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_providers.list_providers(
+ workspace="workspace",
+ )
+ """
+ _response = self._raw_client.list_providers(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ def create_provider(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ host_url: str,
+ project: typing.Optional[str] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ api_key_secret_name: typing.Optional[str] = OMIT,
+ enabled_models: typing.Optional[typing.Sequence[str]] = OMIT,
+ default_extra_body: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ default_extra_headers: typing.Optional[typing.Dict[str, str]] = OMIT,
+ required_extra_body: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ required_extra_headers: typing.Optional[typing.Dict[str, str]] = OMIT,
+ model_deployment_id: typing.Optional[str] = OMIT,
+ status: typing.Optional[ModelProviderStatus] = OMIT,
+ status_message: typing.Optional[str] = OMIT,
+ auth_header_format: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelProvider:
+ """
+ Create a new model provider.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Name of the model provider. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ host_url : str
+ The network endpoint URL for the model provider
+
+ project : typing.Optional[str]
+ The URN of the project associated with this model provider
+
+ description : typing.Optional[str]
+ Optional description of the model provider
+
+ api_key_secret_name : typing.Optional[str]
+ Reference to an API key secret stored in the Secrets service. Create the secret first via secrets API, then pass the secret name here.
+
+ enabled_models : typing.Optional[typing.Sequence[str]]
+ Optional list of specific models to enable from this provider
+
+ default_extra_body : typing.Optional[typing.Dict[str, typing.Any]]
+ Default body parameters for inference requests. Can be overridden by user requests.
+
+ default_extra_headers : typing.Optional[typing.Dict[str, str]]
+ Default headers for inference requests. Can be overridden by user requests.
+
+ required_extra_body : typing.Optional[typing.Dict[str, typing.Any]]
+ Required body parameters for inference requests. Cannot be overridden by user requests.
+
+ required_extra_headers : typing.Optional[typing.Dict[str, str]]
+ Required headers for inference requests. Cannot be overridden by user requests.
+
+ model_deployment_id : typing.Optional[str]
+ Optional reference to the ModelDeployment ID if this provider is being auto-created for a deployment
+
+ status : typing.Optional[ModelProviderStatus]
+ Status of the model provider
+
+ status_message : typing.Optional[str]
+ Status message
+
+ auth_header_format : typing.Optional[str]
+ Jinja2 template string controlling how the API key secret is sent to the upstream. Must contain exactly one variable named `auth_secret`, which is substituted with the resolved secret value at request time. Example: `'X-Api-Key: {{ auth_secret }}'`. If not set, defaults to `'Authorization: Bearer {{ auth_secret }}'`.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelProvider
+ Create a new model provider
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_providers.create_provider(
+ workspace="workspace",
+ name="my-nim-provider",
+ host_url="host_url",
+ )
+ """
+ _response = self._raw_client.create_provider(
+ workspace,
+ name=name,
+ host_url=host_url,
+ project=project,
+ description=description,
+ api_key_secret_name=api_key_secret_name,
+ enabled_models=enabled_models,
+ default_extra_body=default_extra_body,
+ default_extra_headers=default_extra_headers,
+ required_extra_body=required_extra_body,
+ required_extra_headers=required_extra_headers,
+ model_deployment_id=model_deployment_id,
+ status=status,
+ status_message=status_message,
+ auth_header_format=auth_header_format,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def get_provider(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> ModelProvider:
+ """
+ Get a model provider by workspace and name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelProvider
+ Return model provider details
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_providers.get_provider(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.get_provider(workspace, name, request_options=request_options)
+ return _response.data
+
+ def upsert_provider(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ host_url: str,
+ project: typing.Optional[str] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ api_key_secret_name: typing.Optional[str] = OMIT,
+ enabled_models: typing.Optional[typing.Sequence[str]] = OMIT,
+ default_extra_body: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ default_extra_headers: typing.Optional[typing.Dict[str, str]] = OMIT,
+ required_extra_body: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ required_extra_headers: typing.Optional[typing.Dict[str, str]] = OMIT,
+ model_deployment_id: typing.Optional[str] = OMIT,
+ status: typing.Optional[ModelProviderStatus] = OMIT,
+ status_message: typing.Optional[str] = OMIT,
+ auth_header_format: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelProvider:
+ """
+ Create or update a model provider.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ host_url : str
+ The network endpoint URL for the model provider
+
+ project : typing.Optional[str]
+ The URN of the project associated with this model provider
+
+ description : typing.Optional[str]
+ Optional description of the model provider
+
+ api_key_secret_name : typing.Optional[str]
+ Reference to an API key secret stored in the Secrets service. Create the secret first via secrets API, then pass the secret name here.
+
+ enabled_models : typing.Optional[typing.Sequence[str]]
+ Optional list of specific models to enable from this provider
+
+ default_extra_body : typing.Optional[typing.Dict[str, typing.Any]]
+ Default body parameters for inference requests. Can be overridden by user requests.
+
+ default_extra_headers : typing.Optional[typing.Dict[str, str]]
+ Default headers for inference requests. Can be overridden by user requests.
+
+ required_extra_body : typing.Optional[typing.Dict[str, typing.Any]]
+ Required body parameters for inference requests. Cannot be overridden by user requests.
+
+ required_extra_headers : typing.Optional[typing.Dict[str, str]]
+ Required headers for inference requests. Cannot be overridden by user requests.
+
+ model_deployment_id : typing.Optional[str]
+ Optional reference to the ModelDeployment ID if this provider is associated with a deployment
+
+ status : typing.Optional[ModelProviderStatus]
+ Status of the model provider
+
+ status_message : typing.Optional[str]
+ Status message
+
+ auth_header_format : typing.Optional[str]
+ Jinja2 template string controlling how the API key secret is sent to the upstream. Must contain exactly one variable named `auth_secret`, which is substituted with the resolved secret value at request time. Example: `'X-Api-Key: {{ auth_secret }}'`. If not set, defaults to `'Authorization: Bearer {{ auth_secret }}'`.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelProvider
+ Create or update a model provider
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_providers.upsert_provider(
+ workspace="workspace",
+ name="name",
+ host_url="host_url",
+ )
+ """
+ _response = self._raw_client.upsert_provider(
+ workspace,
+ name,
+ host_url=host_url,
+ project=project,
+ description=description,
+ api_key_secret_name=api_key_secret_name,
+ enabled_models=enabled_models,
+ default_extra_body=default_extra_body,
+ default_extra_headers=default_extra_headers,
+ required_extra_body=required_extra_body,
+ required_extra_headers=required_extra_headers,
+ model_deployment_id=model_deployment_id,
+ status=status,
+ status_message=status_message,
+ auth_header_format=auth_header_format,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def delete_provider(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> None:
+ """
+ Delete a model provider by workspace and name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_providers.delete_provider(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.delete_provider(workspace, name, request_options=request_options)
+ return _response.data
+
+ def update_provider_status(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ model_deployment_id: typing.Optional[str] = OMIT,
+ served_models: typing.Optional[typing.Sequence[ServedModelMapping]] = OMIT,
+ status: typing.Optional[ModelProviderStatus] = OMIT,
+ status_message: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelProvider:
+ """
+ Update status-related fields of a model provider.
+
+ This endpoint supports partial updates for fields managed by Models Controller:
+ - model_deployment_id
+ - served_models
+ - status
+ - status_message
+
+ If status is provided without status_message, status_message will be set to empty string.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ model_deployment_id : typing.Optional[str]
+ Reference to the ModelDeployment ID if this provider is associated with a deployment
+
+ served_models : typing.Optional[typing.Sequence[ServedModelMapping]]
+ List of models served by this provider with routing information for IGW
+
+ status : typing.Optional[ModelProviderStatus]
+ Status of the model provider
+
+ status_message : typing.Optional[str]
+ Status message. If status is provided without status_message, defaults to empty string.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelProvider
+ Update status-related fields of a model provider
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.model_providers.update_provider_status(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.update_provider_status(
+ workspace,
+ name,
+ model_deployment_id=model_deployment_id,
+ served_models=served_models,
+ status=status,
+ status_message=status_message,
+ request_options=request_options,
+ )
+ return _response.data
+
+
+class AsyncModelProvidersClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._raw_client = AsyncRawModelProvidersClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> AsyncRawModelProvidersClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ AsyncRawModelProvidersClient
+ """
+ return self._raw_client
+
+ async def list_providers(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[ModelProviderSort] = None,
+ filter: typing.Optional[ModelProviderFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelProvidersPage:
+ """
+ List model providers for a specific workspace.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[ModelProviderSort]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[ModelProviderFilter]
+ Filter model providers by workspace, project, status, model_deployment_id, name, description, host_url, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelProvidersPage
+ Return model providers for a workspace
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_providers.list_providers(
+ workspace="workspace",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_providers(
+ workspace, page=page, page_size=page_size, sort=sort, filter=filter, request_options=request_options
+ )
+ return _response.data
+
+ async def create_provider(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ host_url: str,
+ project: typing.Optional[str] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ api_key_secret_name: typing.Optional[str] = OMIT,
+ enabled_models: typing.Optional[typing.Sequence[str]] = OMIT,
+ default_extra_body: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ default_extra_headers: typing.Optional[typing.Dict[str, str]] = OMIT,
+ required_extra_body: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ required_extra_headers: typing.Optional[typing.Dict[str, str]] = OMIT,
+ model_deployment_id: typing.Optional[str] = OMIT,
+ status: typing.Optional[ModelProviderStatus] = OMIT,
+ status_message: typing.Optional[str] = OMIT,
+ auth_header_format: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelProvider:
+ """
+ Create a new model provider.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Name of the model provider. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ host_url : str
+ The network endpoint URL for the model provider
+
+ project : typing.Optional[str]
+ The URN of the project associated with this model provider
+
+ description : typing.Optional[str]
+ Optional description of the model provider
+
+ api_key_secret_name : typing.Optional[str]
+ Reference to an API key secret stored in the Secrets service. Create the secret first via secrets API, then pass the secret name here.
+
+ enabled_models : typing.Optional[typing.Sequence[str]]
+ Optional list of specific models to enable from this provider
+
+ default_extra_body : typing.Optional[typing.Dict[str, typing.Any]]
+ Default body parameters for inference requests. Can be overridden by user requests.
+
+ default_extra_headers : typing.Optional[typing.Dict[str, str]]
+ Default headers for inference requests. Can be overridden by user requests.
+
+ required_extra_body : typing.Optional[typing.Dict[str, typing.Any]]
+ Required body parameters for inference requests. Cannot be overridden by user requests.
+
+ required_extra_headers : typing.Optional[typing.Dict[str, str]]
+ Required headers for inference requests. Cannot be overridden by user requests.
+
+ model_deployment_id : typing.Optional[str]
+ Optional reference to the ModelDeployment ID if this provider is being auto-created for a deployment
+
+ status : typing.Optional[ModelProviderStatus]
+ Status of the model provider
+
+ status_message : typing.Optional[str]
+ Status message
+
+ auth_header_format : typing.Optional[str]
+ Jinja2 template string controlling how the API key secret is sent to the upstream. Must contain exactly one variable named `auth_secret`, which is substituted with the resolved secret value at request time. Example: `'X-Api-Key: {{ auth_secret }}'`. If not set, defaults to `'Authorization: Bearer {{ auth_secret }}'`.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelProvider
+ Create a new model provider
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_providers.create_provider(
+ workspace="workspace",
+ name="my-nim-provider",
+ host_url="host_url",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.create_provider(
+ workspace,
+ name=name,
+ host_url=host_url,
+ project=project,
+ description=description,
+ api_key_secret_name=api_key_secret_name,
+ enabled_models=enabled_models,
+ default_extra_body=default_extra_body,
+ default_extra_headers=default_extra_headers,
+ required_extra_body=required_extra_body,
+ required_extra_headers=required_extra_headers,
+ model_deployment_id=model_deployment_id,
+ status=status,
+ status_message=status_message,
+ auth_header_format=auth_header_format,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def get_provider(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> ModelProvider:
+ """
+ Get a model provider by workspace and name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelProvider
+ Return model provider details
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_providers.get_provider(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_provider(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def upsert_provider(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ host_url: str,
+ project: typing.Optional[str] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ api_key_secret_name: typing.Optional[str] = OMIT,
+ enabled_models: typing.Optional[typing.Sequence[str]] = OMIT,
+ default_extra_body: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ default_extra_headers: typing.Optional[typing.Dict[str, str]] = OMIT,
+ required_extra_body: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ required_extra_headers: typing.Optional[typing.Dict[str, str]] = OMIT,
+ model_deployment_id: typing.Optional[str] = OMIT,
+ status: typing.Optional[ModelProviderStatus] = OMIT,
+ status_message: typing.Optional[str] = OMIT,
+ auth_header_format: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelProvider:
+ """
+ Create or update a model provider.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ host_url : str
+ The network endpoint URL for the model provider
+
+ project : typing.Optional[str]
+ The URN of the project associated with this model provider
+
+ description : typing.Optional[str]
+ Optional description of the model provider
+
+ api_key_secret_name : typing.Optional[str]
+ Reference to an API key secret stored in the Secrets service. Create the secret first via secrets API, then pass the secret name here.
+
+ enabled_models : typing.Optional[typing.Sequence[str]]
+ Optional list of specific models to enable from this provider
+
+ default_extra_body : typing.Optional[typing.Dict[str, typing.Any]]
+ Default body parameters for inference requests. Can be overridden by user requests.
+
+ default_extra_headers : typing.Optional[typing.Dict[str, str]]
+ Default headers for inference requests. Can be overridden by user requests.
+
+ required_extra_body : typing.Optional[typing.Dict[str, typing.Any]]
+ Required body parameters for inference requests. Cannot be overridden by user requests.
+
+ required_extra_headers : typing.Optional[typing.Dict[str, str]]
+ Required headers for inference requests. Cannot be overridden by user requests.
+
+ model_deployment_id : typing.Optional[str]
+ Optional reference to the ModelDeployment ID if this provider is associated with a deployment
+
+ status : typing.Optional[ModelProviderStatus]
+ Status of the model provider
+
+ status_message : typing.Optional[str]
+ Status message
+
+ auth_header_format : typing.Optional[str]
+ Jinja2 template string controlling how the API key secret is sent to the upstream. Must contain exactly one variable named `auth_secret`, which is substituted with the resolved secret value at request time. Example: `'X-Api-Key: {{ auth_secret }}'`. If not set, defaults to `'Authorization: Bearer {{ auth_secret }}'`.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelProvider
+ Create or update a model provider
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_providers.upsert_provider(
+ workspace="workspace",
+ name="name",
+ host_url="host_url",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.upsert_provider(
+ workspace,
+ name,
+ host_url=host_url,
+ project=project,
+ description=description,
+ api_key_secret_name=api_key_secret_name,
+ enabled_models=enabled_models,
+ default_extra_body=default_extra_body,
+ default_extra_headers=default_extra_headers,
+ required_extra_body=required_extra_body,
+ required_extra_headers=required_extra_headers,
+ model_deployment_id=model_deployment_id,
+ status=status,
+ status_message=status_message,
+ auth_header_format=auth_header_format,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def delete_provider(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> None:
+ """
+ Delete a model provider by workspace and name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_providers.delete_provider(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.delete_provider(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def update_provider_status(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ model_deployment_id: typing.Optional[str] = OMIT,
+ served_models: typing.Optional[typing.Sequence[ServedModelMapping]] = OMIT,
+ status: typing.Optional[ModelProviderStatus] = OMIT,
+ status_message: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelProvider:
+ """
+ Update status-related fields of a model provider.
+
+ This endpoint supports partial updates for fields managed by Models Controller:
+ - model_deployment_id
+ - served_models
+ - status
+ - status_message
+
+ If status is provided without status_message, status_message will be set to empty string.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ model_deployment_id : typing.Optional[str]
+ Reference to the ModelDeployment ID if this provider is associated with a deployment
+
+ served_models : typing.Optional[typing.Sequence[ServedModelMapping]]
+ List of models served by this provider with routing information for IGW
+
+ status : typing.Optional[ModelProviderStatus]
+ Status of the model provider
+
+ status_message : typing.Optional[str]
+ Status message. If status is provided without status_message, defaults to empty string.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelProvider
+ Update status-related fields of a model provider
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.model_providers.update_provider_status(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.update_provider_status(
+ workspace,
+ name,
+ model_deployment_id=model_deployment_id,
+ served_models=served_models,
+ status=status,
+ status_message=status_message,
+ request_options=request_options,
+ )
+ return _response.data
diff --git a/sdks/python/model_providers/raw_client.py b/sdks/python/model_providers/raw_client.py
new file mode 100644
index 0000000000..769a68f44a
--- /dev/null
+++ b/sdks/python/model_providers/raw_client.py
@@ -0,0 +1,1112 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+from json.decoder import JSONDecodeError
+
+from ..core.api_error import ApiError
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.http_response import AsyncHttpResponse, HttpResponse
+from ..core.jsonable_encoder import encode_path_param
+from ..core.parse_error import ParsingError
+from ..core.pydantic_utilities import parse_obj_as
+from ..core.request_options import RequestOptions
+from ..core.serialization import convert_and_respect_annotation_metadata
+from ..errors.unprocessable_entity_error import UnprocessableEntityError
+from ..types.model_provider import ModelProvider
+from ..types.model_provider_filter import ModelProviderFilter
+from ..types.model_provider_sort import ModelProviderSort
+from ..types.model_provider_status import ModelProviderStatus
+from ..types.model_providers_page import ModelProvidersPage
+from ..types.served_model_mapping import ServedModelMapping
+from pydantic import ValidationError
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class RawModelProvidersClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ def list_providers(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[ModelProviderSort] = None,
+ filter: typing.Optional[ModelProviderFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ModelProvidersPage]:
+ """
+ List model providers for a specific workspace.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[ModelProviderSort]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[ModelProviderFilter]
+ Filter model providers by workspace, project, status, model_deployment_id, name, description, host_url, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ModelProvidersPage]
+ Return model providers for a workspace
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/providers",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=ModelProviderFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelProvidersPage,
+ parse_obj_as(
+ type_=ModelProvidersPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def create_provider(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ host_url: str,
+ project: typing.Optional[str] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ api_key_secret_name: typing.Optional[str] = OMIT,
+ enabled_models: typing.Optional[typing.Sequence[str]] = OMIT,
+ default_extra_body: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ default_extra_headers: typing.Optional[typing.Dict[str, str]] = OMIT,
+ required_extra_body: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ required_extra_headers: typing.Optional[typing.Dict[str, str]] = OMIT,
+ model_deployment_id: typing.Optional[str] = OMIT,
+ status: typing.Optional[ModelProviderStatus] = OMIT,
+ status_message: typing.Optional[str] = OMIT,
+ auth_header_format: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ModelProvider]:
+ """
+ Create a new model provider.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Name of the model provider. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ host_url : str
+ The network endpoint URL for the model provider
+
+ project : typing.Optional[str]
+ The URN of the project associated with this model provider
+
+ description : typing.Optional[str]
+ Optional description of the model provider
+
+ api_key_secret_name : typing.Optional[str]
+ Reference to an API key secret stored in the Secrets service. Create the secret first via secrets API, then pass the secret name here.
+
+ enabled_models : typing.Optional[typing.Sequence[str]]
+ Optional list of specific models to enable from this provider
+
+ default_extra_body : typing.Optional[typing.Dict[str, typing.Any]]
+ Default body parameters for inference requests. Can be overridden by user requests.
+
+ default_extra_headers : typing.Optional[typing.Dict[str, str]]
+ Default headers for inference requests. Can be overridden by user requests.
+
+ required_extra_body : typing.Optional[typing.Dict[str, typing.Any]]
+ Required body parameters for inference requests. Cannot be overridden by user requests.
+
+ required_extra_headers : typing.Optional[typing.Dict[str, str]]
+ Required headers for inference requests. Cannot be overridden by user requests.
+
+ model_deployment_id : typing.Optional[str]
+ Optional reference to the ModelDeployment ID if this provider is being auto-created for a deployment
+
+ status : typing.Optional[ModelProviderStatus]
+ Status of the model provider
+
+ status_message : typing.Optional[str]
+ Status message
+
+ auth_header_format : typing.Optional[str]
+ Jinja2 template string controlling how the API key secret is sent to the upstream. Must contain exactly one variable named `auth_secret`, which is substituted with the resolved secret value at request time. Example: `'X-Api-Key: {{ auth_secret }}'`. If not set, defaults to `'Authorization: Bearer {{ auth_secret }}'`.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ModelProvider]
+ Create a new model provider
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/providers",
+ method="POST",
+ json={
+ "name": name,
+ "project": project,
+ "description": description,
+ "host_url": host_url,
+ "api_key_secret_name": api_key_secret_name,
+ "enabled_models": enabled_models,
+ "default_extra_body": default_extra_body,
+ "default_extra_headers": default_extra_headers,
+ "required_extra_body": required_extra_body,
+ "required_extra_headers": required_extra_headers,
+ "model_deployment_id": model_deployment_id,
+ "status": status,
+ "status_message": status_message,
+ "auth_header_format": auth_header_format,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelProvider,
+ parse_obj_as(
+ type_=ModelProvider, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_provider(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[ModelProvider]:
+ """
+ Get a model provider by workspace and name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ModelProvider]
+ Return model provider details
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/providers/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelProvider,
+ parse_obj_as(
+ type_=ModelProvider, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def upsert_provider(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ host_url: str,
+ project: typing.Optional[str] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ api_key_secret_name: typing.Optional[str] = OMIT,
+ enabled_models: typing.Optional[typing.Sequence[str]] = OMIT,
+ default_extra_body: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ default_extra_headers: typing.Optional[typing.Dict[str, str]] = OMIT,
+ required_extra_body: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ required_extra_headers: typing.Optional[typing.Dict[str, str]] = OMIT,
+ model_deployment_id: typing.Optional[str] = OMIT,
+ status: typing.Optional[ModelProviderStatus] = OMIT,
+ status_message: typing.Optional[str] = OMIT,
+ auth_header_format: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ModelProvider]:
+ """
+ Create or update a model provider.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ host_url : str
+ The network endpoint URL for the model provider
+
+ project : typing.Optional[str]
+ The URN of the project associated with this model provider
+
+ description : typing.Optional[str]
+ Optional description of the model provider
+
+ api_key_secret_name : typing.Optional[str]
+ Reference to an API key secret stored in the Secrets service. Create the secret first via secrets API, then pass the secret name here.
+
+ enabled_models : typing.Optional[typing.Sequence[str]]
+ Optional list of specific models to enable from this provider
+
+ default_extra_body : typing.Optional[typing.Dict[str, typing.Any]]
+ Default body parameters for inference requests. Can be overridden by user requests.
+
+ default_extra_headers : typing.Optional[typing.Dict[str, str]]
+ Default headers for inference requests. Can be overridden by user requests.
+
+ required_extra_body : typing.Optional[typing.Dict[str, typing.Any]]
+ Required body parameters for inference requests. Cannot be overridden by user requests.
+
+ required_extra_headers : typing.Optional[typing.Dict[str, str]]
+ Required headers for inference requests. Cannot be overridden by user requests.
+
+ model_deployment_id : typing.Optional[str]
+ Optional reference to the ModelDeployment ID if this provider is associated with a deployment
+
+ status : typing.Optional[ModelProviderStatus]
+ Status of the model provider
+
+ status_message : typing.Optional[str]
+ Status message
+
+ auth_header_format : typing.Optional[str]
+ Jinja2 template string controlling how the API key secret is sent to the upstream. Must contain exactly one variable named `auth_secret`, which is substituted with the resolved secret value at request time. Example: `'X-Api-Key: {{ auth_secret }}'`. If not set, defaults to `'Authorization: Bearer {{ auth_secret }}'`.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ModelProvider]
+ Create or update a model provider
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/providers/{encode_path_param(name)}",
+ method="PUT",
+ json={
+ "project": project,
+ "description": description,
+ "host_url": host_url,
+ "api_key_secret_name": api_key_secret_name,
+ "enabled_models": enabled_models,
+ "default_extra_body": default_extra_body,
+ "default_extra_headers": default_extra_headers,
+ "required_extra_body": required_extra_body,
+ "required_extra_headers": required_extra_headers,
+ "model_deployment_id": model_deployment_id,
+ "status": status,
+ "status_message": status_message,
+ "auth_header_format": auth_header_format,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelProvider,
+ parse_obj_as(
+ type_=ModelProvider, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def delete_provider(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[None]:
+ """
+ Delete a model provider by workspace and name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[None]
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/providers/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return HttpResponse(response=_response, data=None)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def update_provider_status(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ model_deployment_id: typing.Optional[str] = OMIT,
+ served_models: typing.Optional[typing.Sequence[ServedModelMapping]] = OMIT,
+ status: typing.Optional[ModelProviderStatus] = OMIT,
+ status_message: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ModelProvider]:
+ """
+ Update status-related fields of a model provider.
+
+ This endpoint supports partial updates for fields managed by Models Controller:
+ - model_deployment_id
+ - served_models
+ - status
+ - status_message
+
+ If status is provided without status_message, status_message will be set to empty string.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ model_deployment_id : typing.Optional[str]
+ Reference to the ModelDeployment ID if this provider is associated with a deployment
+
+ served_models : typing.Optional[typing.Sequence[ServedModelMapping]]
+ List of models served by this provider with routing information for IGW
+
+ status : typing.Optional[ModelProviderStatus]
+ Status of the model provider
+
+ status_message : typing.Optional[str]
+ Status message. If status is provided without status_message, defaults to empty string.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ModelProvider]
+ Update status-related fields of a model provider
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/providers/{encode_path_param(name)}/status",
+ method="PUT",
+ json={
+ "model_deployment_id": model_deployment_id,
+ "served_models": convert_and_respect_annotation_metadata(
+ object_=served_models, annotation=typing.Sequence[ServedModelMapping], direction="write"
+ ),
+ "status": status,
+ "status_message": status_message,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelProvider,
+ parse_obj_as(
+ type_=ModelProvider, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+
+class AsyncRawModelProvidersClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ async def list_providers(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[ModelProviderSort] = None,
+ filter: typing.Optional[ModelProviderFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ModelProvidersPage]:
+ """
+ List model providers for a specific workspace.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[ModelProviderSort]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ filter : typing.Optional[ModelProviderFilter]
+ Filter model providers by workspace, project, status, model_deployment_id, name, description, host_url, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ModelProvidersPage]
+ Return model providers for a workspace
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/providers",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=ModelProviderFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelProvidersPage,
+ parse_obj_as(
+ type_=ModelProvidersPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def create_provider(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ host_url: str,
+ project: typing.Optional[str] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ api_key_secret_name: typing.Optional[str] = OMIT,
+ enabled_models: typing.Optional[typing.Sequence[str]] = OMIT,
+ default_extra_body: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ default_extra_headers: typing.Optional[typing.Dict[str, str]] = OMIT,
+ required_extra_body: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ required_extra_headers: typing.Optional[typing.Dict[str, str]] = OMIT,
+ model_deployment_id: typing.Optional[str] = OMIT,
+ status: typing.Optional[ModelProviderStatus] = OMIT,
+ status_message: typing.Optional[str] = OMIT,
+ auth_header_format: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ModelProvider]:
+ """
+ Create a new model provider.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Name of the model provider. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ host_url : str
+ The network endpoint URL for the model provider
+
+ project : typing.Optional[str]
+ The URN of the project associated with this model provider
+
+ description : typing.Optional[str]
+ Optional description of the model provider
+
+ api_key_secret_name : typing.Optional[str]
+ Reference to an API key secret stored in the Secrets service. Create the secret first via secrets API, then pass the secret name here.
+
+ enabled_models : typing.Optional[typing.Sequence[str]]
+ Optional list of specific models to enable from this provider
+
+ default_extra_body : typing.Optional[typing.Dict[str, typing.Any]]
+ Default body parameters for inference requests. Can be overridden by user requests.
+
+ default_extra_headers : typing.Optional[typing.Dict[str, str]]
+ Default headers for inference requests. Can be overridden by user requests.
+
+ required_extra_body : typing.Optional[typing.Dict[str, typing.Any]]
+ Required body parameters for inference requests. Cannot be overridden by user requests.
+
+ required_extra_headers : typing.Optional[typing.Dict[str, str]]
+ Required headers for inference requests. Cannot be overridden by user requests.
+
+ model_deployment_id : typing.Optional[str]
+ Optional reference to the ModelDeployment ID if this provider is being auto-created for a deployment
+
+ status : typing.Optional[ModelProviderStatus]
+ Status of the model provider
+
+ status_message : typing.Optional[str]
+ Status message
+
+ auth_header_format : typing.Optional[str]
+ Jinja2 template string controlling how the API key secret is sent to the upstream. Must contain exactly one variable named `auth_secret`, which is substituted with the resolved secret value at request time. Example: `'X-Api-Key: {{ auth_secret }}'`. If not set, defaults to `'Authorization: Bearer {{ auth_secret }}'`.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ModelProvider]
+ Create a new model provider
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/providers",
+ method="POST",
+ json={
+ "name": name,
+ "project": project,
+ "description": description,
+ "host_url": host_url,
+ "api_key_secret_name": api_key_secret_name,
+ "enabled_models": enabled_models,
+ "default_extra_body": default_extra_body,
+ "default_extra_headers": default_extra_headers,
+ "required_extra_body": required_extra_body,
+ "required_extra_headers": required_extra_headers,
+ "model_deployment_id": model_deployment_id,
+ "status": status,
+ "status_message": status_message,
+ "auth_header_format": auth_header_format,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelProvider,
+ parse_obj_as(
+ type_=ModelProvider, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_provider(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[ModelProvider]:
+ """
+ Get a model provider by workspace and name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ModelProvider]
+ Return model provider details
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/providers/{encode_path_param(name)}",
+ method="GET",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelProvider,
+ parse_obj_as(
+ type_=ModelProvider, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def upsert_provider(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ host_url: str,
+ project: typing.Optional[str] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ api_key_secret_name: typing.Optional[str] = OMIT,
+ enabled_models: typing.Optional[typing.Sequence[str]] = OMIT,
+ default_extra_body: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ default_extra_headers: typing.Optional[typing.Dict[str, str]] = OMIT,
+ required_extra_body: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ required_extra_headers: typing.Optional[typing.Dict[str, str]] = OMIT,
+ model_deployment_id: typing.Optional[str] = OMIT,
+ status: typing.Optional[ModelProviderStatus] = OMIT,
+ status_message: typing.Optional[str] = OMIT,
+ auth_header_format: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ModelProvider]:
+ """
+ Create or update a model provider.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ host_url : str
+ The network endpoint URL for the model provider
+
+ project : typing.Optional[str]
+ The URN of the project associated with this model provider
+
+ description : typing.Optional[str]
+ Optional description of the model provider
+
+ api_key_secret_name : typing.Optional[str]
+ Reference to an API key secret stored in the Secrets service. Create the secret first via secrets API, then pass the secret name here.
+
+ enabled_models : typing.Optional[typing.Sequence[str]]
+ Optional list of specific models to enable from this provider
+
+ default_extra_body : typing.Optional[typing.Dict[str, typing.Any]]
+ Default body parameters for inference requests. Can be overridden by user requests.
+
+ default_extra_headers : typing.Optional[typing.Dict[str, str]]
+ Default headers for inference requests. Can be overridden by user requests.
+
+ required_extra_body : typing.Optional[typing.Dict[str, typing.Any]]
+ Required body parameters for inference requests. Cannot be overridden by user requests.
+
+ required_extra_headers : typing.Optional[typing.Dict[str, str]]
+ Required headers for inference requests. Cannot be overridden by user requests.
+
+ model_deployment_id : typing.Optional[str]
+ Optional reference to the ModelDeployment ID if this provider is associated with a deployment
+
+ status : typing.Optional[ModelProviderStatus]
+ Status of the model provider
+
+ status_message : typing.Optional[str]
+ Status message
+
+ auth_header_format : typing.Optional[str]
+ Jinja2 template string controlling how the API key secret is sent to the upstream. Must contain exactly one variable named `auth_secret`, which is substituted with the resolved secret value at request time. Example: `'X-Api-Key: {{ auth_secret }}'`. If not set, defaults to `'Authorization: Bearer {{ auth_secret }}'`.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ModelProvider]
+ Create or update a model provider
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/providers/{encode_path_param(name)}",
+ method="PUT",
+ json={
+ "project": project,
+ "description": description,
+ "host_url": host_url,
+ "api_key_secret_name": api_key_secret_name,
+ "enabled_models": enabled_models,
+ "default_extra_body": default_extra_body,
+ "default_extra_headers": default_extra_headers,
+ "required_extra_body": required_extra_body,
+ "required_extra_headers": required_extra_headers,
+ "model_deployment_id": model_deployment_id,
+ "status": status,
+ "status_message": status_message,
+ "auth_header_format": auth_header_format,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelProvider,
+ parse_obj_as(
+ type_=ModelProvider, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def delete_provider(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[None]:
+ """
+ Delete a model provider by workspace and name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[None]
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/providers/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return AsyncHttpResponse(response=_response, data=None)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def update_provider_status(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ model_deployment_id: typing.Optional[str] = OMIT,
+ served_models: typing.Optional[typing.Sequence[ServedModelMapping]] = OMIT,
+ status: typing.Optional[ModelProviderStatus] = OMIT,
+ status_message: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ModelProvider]:
+ """
+ Update status-related fields of a model provider.
+
+ This endpoint supports partial updates for fields managed by Models Controller:
+ - model_deployment_id
+ - served_models
+ - status
+ - status_message
+
+ If status is provided without status_message, status_message will be set to empty string.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ model_deployment_id : typing.Optional[str]
+ Reference to the ModelDeployment ID if this provider is associated with a deployment
+
+ served_models : typing.Optional[typing.Sequence[ServedModelMapping]]
+ List of models served by this provider with routing information for IGW
+
+ status : typing.Optional[ModelProviderStatus]
+ Status of the model provider
+
+ status_message : typing.Optional[str]
+ Status message. If status is provided without status_message, defaults to empty string.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ModelProvider]
+ Update status-related fields of a model provider
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/providers/{encode_path_param(name)}/status",
+ method="PUT",
+ json={
+ "model_deployment_id": model_deployment_id,
+ "served_models": convert_and_respect_annotation_metadata(
+ object_=served_models, annotation=typing.Sequence[ServedModelMapping], direction="write"
+ ),
+ "status": status,
+ "status_message": status_message,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelProvider,
+ parse_obj_as(
+ type_=ModelProvider, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
diff --git a/sdks/python/models/__init__.py b/sdks/python/models/__init__.py
new file mode 100644
index 0000000000..5cde0202dc
--- /dev/null
+++ b/sdks/python/models/__init__.py
@@ -0,0 +1,4 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
diff --git a/sdks/python/models/client.py b/sdks/python/models/client.py
new file mode 100644
index 0000000000..1a19f738e5
--- /dev/null
+++ b/sdks/python/models/client.py
@@ -0,0 +1,1237 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.request_options import RequestOptions
+from ..types.adapter import Adapter
+from ..types.api_endpoint_data import ApiEndpointData
+from ..types.backend_format import BackendFormat
+from ..types.finetuning_type import FinetuningType
+from ..types.lora import Lora
+from ..types.model_entity import ModelEntity
+from ..types.model_entity_filter import ModelEntityFilter
+from ..types.model_entity_sort_field import ModelEntitySortField
+from ..types.model_entitys_page import ModelEntitysPage
+from ..types.model_spec import ModelSpec
+from ..types.prompt_data import PromptData
+from .raw_client import AsyncRawModelsClient, RawModelsClient
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class ModelsClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._raw_client = RawModelsClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> RawModelsClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ RawModelsClient
+ """
+ return self._raw_client
+
+ def list_models(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[ModelEntitySortField] = None,
+ verbose: typing.Optional[bool] = None,
+ filter: typing.Optional[ModelEntityFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelEntitysPage:
+ """
+ List Models endpoint with filtering, pagination, and sorting.
+
+ Supports filter parameters for various criteria (including peft, custom fields),
+ pagination (page, page_size), sorting, and workspace filtering via query parameter.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[ModelEntitySortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ verbose : typing.Optional[bool]
+ Whether to include full spec details
+
+ filter : typing.Optional[ModelEntityFilter]
+ Filter models by name, project, workspace, base_model, adapters, finetuning_type, prompt, lora_enabled, description, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelEntitysPage
+ Return a list of models
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.models.list_models(
+ workspace="workspace",
+ )
+ """
+ _response = self._raw_client.list_models(
+ workspace,
+ page=page,
+ page_size=page_size,
+ sort=sort,
+ verbose=verbose,
+ filter=filter,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def create_model(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ project: typing.Optional[str] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ spec: typing.Optional[ModelSpec] = OMIT,
+ finetuning_type: typing.Optional[FinetuningType] = OMIT,
+ fileset: typing.Optional[str] = OMIT,
+ base_model: typing.Optional[str] = OMIT,
+ api_endpoint: typing.Optional[ApiEndpointData] = OMIT,
+ backend_format: typing.Optional[BackendFormat] = OMIT,
+ prompt: typing.Optional[PromptData] = OMIT,
+ custom_fields: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ ownership: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ model_providers: typing.Optional[typing.Sequence[str]] = OMIT,
+ trust_remote_code: typing.Optional[bool] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelEntity:
+ """
+ Create a new model entity.
+
+ This endpoint creates a new Model Entity in the Models service database.
+ The Model Entity will be registered for use within the platform.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Name of the model entity. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ project : typing.Optional[str]
+ The URN of the project associated with this model entity
+
+ description : typing.Optional[str]
+ Optional description of the model
+
+ spec : typing.Optional[ModelSpec]
+ Detailed specification for the model - Automatically generated by the platform at creation when fileset provided.
+
+ finetuning_type : typing.Optional[FinetuningType]
+ Set for full weight finetuned models
+
+ fileset : typing.Optional[str]
+ A set of checkpoint files, configs, and other auxiliary info associated with this model - expected format {workspace}/{fileset_name}
+
+ base_model : typing.Optional[str]
+ Link to another model which is used as a base for the current model
+
+ api_endpoint : typing.Optional[ApiEndpointData]
+ Data about the inference endpoint for this model
+
+ backend_format : typing.Optional[BackendFormat]
+ Inference API wire format expected by the backend. If unset, inference routing treats the model as OPENAI_CHAT.
+
+ prompt : typing.Optional[PromptData]
+ Configuration for prompt engineering
+
+ custom_fields : typing.Optional[typing.Dict[str, typing.Any]]
+ Custom fields for additional metadata
+
+ ownership : typing.Optional[typing.Dict[str, typing.Any]]
+ Ownership information for the model
+
+ model_providers : typing.Optional[typing.Sequence[str]]
+ List of ModelProvider workspace/name resource names that provide inference for this Model Entity
+
+ trust_remote_code : typing.Optional[bool]
+ Whether to trust remote code for the checkpoint.
+ Some models without support in certain libraries such as Transformers require additional custom Python code to execute.
+ Due to security ramifications of running arbitrary code, this can only be set to true on one of the following conditions:
+ (1) the model's fileset's source is pre-approved in the platform config, or
+ (2) the user creating this model is an administrator.
+
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelEntity
+ Create a new model entity
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.models.create_model(
+ workspace="workspace",
+ name="llama-3.1-8b",
+ )
+ """
+ _response = self._raw_client.create_model(
+ workspace,
+ name=name,
+ project=project,
+ description=description,
+ spec=spec,
+ finetuning_type=finetuning_type,
+ fileset=fileset,
+ base_model=base_model,
+ api_endpoint=api_endpoint,
+ backend_format=backend_format,
+ prompt=prompt,
+ custom_fields=custom_fields,
+ ownership=ownership,
+ model_providers=model_providers,
+ trust_remote_code=trust_remote_code,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def create_model_adapter(
+ self,
+ workspace: str,
+ model_name: str,
+ *,
+ name: str,
+ fileset: str,
+ finetuning_type: FinetuningType,
+ description: typing.Optional[str] = OMIT,
+ enabled: typing.Optional[bool] = OMIT,
+ lora_config: typing.Optional[Lora] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Adapter:
+ """
+ Adds an Adapter to the Model
+
+ Parameters
+ ----------
+ workspace : str
+
+ model_name : str
+
+ name : str
+ Name of the adapter. Name must be unique in the workspace. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ fileset : str
+ Location where adapter files are stored - expected format {workspace}/{fileset_name}
+
+ finetuning_type : FinetuningType
+ Type of finetuning (LORA, P_TUNING, etc.)
+
+ description : typing.Optional[str]
+ Optional description of the adapter
+
+ enabled : typing.Optional[bool]
+ Whether to make this adapter available for inference post training
+
+ lora_config : typing.Optional[Lora]
+ Lora configuration specifics
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Adapter
+ Register a new adapter to the model
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.models.create_model_adapter(
+ workspace="workspace",
+ model_name="model_name",
+ name="lora-adapter-v1",
+ fileset="fileset",
+ finetuning_type="lora_merged",
+ )
+ """
+ _response = self._raw_client.create_model_adapter(
+ workspace,
+ model_name,
+ name=name,
+ fileset=fileset,
+ finetuning_type=finetuning_type,
+ description=description,
+ enabled=enabled,
+ lora_config=lora_config,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def delete_model_adapter(
+ self, workspace: str, model_name: str, adapter: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> None:
+ """
+ Delete Adapter from Model entity.
+
+ Permanently deletes an adapter from a model entity, if it was deployed, it will be cleaned up automatically.
+
+ Parameters
+ ----------
+ workspace : str
+
+ model_name : str
+
+ adapter : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.models.delete_model_adapter(
+ workspace="workspace",
+ model_name="model_name",
+ adapter="adapter",
+ )
+ """
+ _response = self._raw_client.delete_model_adapter(
+ workspace, model_name, adapter, request_options=request_options
+ )
+ return _response.data
+
+ def update_model_adapter(
+ self,
+ workspace: str,
+ model_name: str,
+ adapter: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ enabled: typing.Optional[bool] = OMIT,
+ fileset: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Adapter:
+ """
+ Update Adapter deployment or description.
+
+ Parameters
+ ----------
+ workspace : str
+
+ model_name : str
+
+ adapter : str
+
+ description : typing.Optional[str]
+ Optional description of the adapter
+
+ enabled : typing.Optional[bool]
+ Whether to make this adapter available for inference post training
+
+ fileset : typing.Optional[str]
+ Updated fileset for the adapter
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Adapter
+ Update adapter metadata
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.models.update_model_adapter(
+ workspace="workspace",
+ model_name="model_name",
+ adapter="adapter",
+ )
+ """
+ _response = self._raw_client.update_model_adapter(
+ workspace,
+ model_name,
+ adapter,
+ description=description,
+ enabled=enabled,
+ fileset=fileset,
+ request_options=request_options,
+ )
+ return _response.data
+
+ def get_model(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ verbose: typing.Optional[bool] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelEntity:
+ """
+ Get Model by Workspace and Name.
+
+ Returns the details of a specific model entity identified by its workspace and name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ verbose : typing.Optional[bool]
+ Whether to include full spec details
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelEntity
+ Return model details
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.models.get_model(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.get_model(workspace, name, verbose=verbose, request_options=request_options)
+ return _response.data
+
+ def delete_model(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> None:
+ """
+ Delete Model entity.
+
+ Permanently deletes a model entity from the platform.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.models.delete_model(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.delete_model(workspace, name, request_options=request_options)
+ return _response.data
+
+ def update_model(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ verbose: typing.Optional[bool] = None,
+ description: typing.Optional[str] = OMIT,
+ spec: typing.Optional[ModelSpec] = OMIT,
+ fileset: typing.Optional[str] = OMIT,
+ finetuning_type: typing.Optional[FinetuningType] = OMIT,
+ base_model: typing.Optional[str] = OMIT,
+ api_endpoint: typing.Optional[ApiEndpointData] = OMIT,
+ backend_format: typing.Optional[BackendFormat] = OMIT,
+ prompt: typing.Optional[PromptData] = OMIT,
+ custom_fields: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ ownership: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ model_providers: typing.Optional[typing.Sequence[str]] = OMIT,
+ trust_remote_code: typing.Optional[bool] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelEntity:
+ """
+ Update Model metadata.
+
+ Updates the metadata of an existing model entity. If the request body has an empty field,
+ the old value is kept.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ verbose : typing.Optional[bool]
+ Whether to include full spec details
+
+ description : typing.Optional[str]
+ Optional description of the model
+
+ spec : typing.Optional[ModelSpec]
+ Detailed specification for the model
+
+ fileset : typing.Optional[str]
+ A set of checkpoint files, configs, and other auxiliary info associated with this model - expected format {workspace}/{fileset_name}
+
+ finetuning_type : typing.Optional[FinetuningType]
+ Set for full weight finetuned models
+
+ base_model : typing.Optional[str]
+ Link to another model which is used as a base for the current model
+
+ api_endpoint : typing.Optional[ApiEndpointData]
+ Data about the inference endpoint for this model
+
+ backend_format : typing.Optional[BackendFormat]
+ Inference API wire format expected by the backend. If unset, inference routing treats the model as OPENAI_CHAT.
+
+ prompt : typing.Optional[PromptData]
+ Configuration for prompt engineering
+
+ custom_fields : typing.Optional[typing.Dict[str, typing.Any]]
+ Custom fields for additional metadata
+
+ ownership : typing.Optional[typing.Dict[str, typing.Any]]
+ Ownership information for the model
+
+ model_providers : typing.Optional[typing.Sequence[str]]
+ List of ModelProvider workspace/name resource names that provide inference for this Model Entity
+
+ trust_remote_code : typing.Optional[bool]
+ Whether to trust remote code for the checkpoint.
+ Some models without support in certain libraries such as Transformers require additional custom Python code to execute.
+ Due to security ramifications of running arbitrary code, this can only be set to true on one of the following conditions:
+ (1) the model's fileset's source is pre-approved in the platform config, or
+ (2) the user creating this model is an administrator.
+
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelEntity
+ Update model metadata
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.models.update_model(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.update_model(
+ workspace,
+ name,
+ verbose=verbose,
+ description=description,
+ spec=spec,
+ fileset=fileset,
+ finetuning_type=finetuning_type,
+ base_model=base_model,
+ api_endpoint=api_endpoint,
+ backend_format=backend_format,
+ prompt=prompt,
+ custom_fields=custom_fields,
+ ownership=ownership,
+ model_providers=model_providers,
+ trust_remote_code=trust_remote_code,
+ request_options=request_options,
+ )
+ return _response.data
+
+
+class AsyncModelsClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._raw_client = AsyncRawModelsClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> AsyncRawModelsClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ AsyncRawModelsClient
+ """
+ return self._raw_client
+
+ async def list_models(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[ModelEntitySortField] = None,
+ verbose: typing.Optional[bool] = None,
+ filter: typing.Optional[ModelEntityFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelEntitysPage:
+ """
+ List Models endpoint with filtering, pagination, and sorting.
+
+ Supports filter parameters for various criteria (including peft, custom fields),
+ pagination (page, page_size), sorting, and workspace filtering via query parameter.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[ModelEntitySortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ verbose : typing.Optional[bool]
+ Whether to include full spec details
+
+ filter : typing.Optional[ModelEntityFilter]
+ Filter models by name, project, workspace, base_model, adapters, finetuning_type, prompt, lora_enabled, description, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelEntitysPage
+ Return a list of models
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.models.list_models(
+ workspace="workspace",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.list_models(
+ workspace,
+ page=page,
+ page_size=page_size,
+ sort=sort,
+ verbose=verbose,
+ filter=filter,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def create_model(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ project: typing.Optional[str] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ spec: typing.Optional[ModelSpec] = OMIT,
+ finetuning_type: typing.Optional[FinetuningType] = OMIT,
+ fileset: typing.Optional[str] = OMIT,
+ base_model: typing.Optional[str] = OMIT,
+ api_endpoint: typing.Optional[ApiEndpointData] = OMIT,
+ backend_format: typing.Optional[BackendFormat] = OMIT,
+ prompt: typing.Optional[PromptData] = OMIT,
+ custom_fields: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ ownership: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ model_providers: typing.Optional[typing.Sequence[str]] = OMIT,
+ trust_remote_code: typing.Optional[bool] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelEntity:
+ """
+ Create a new model entity.
+
+ This endpoint creates a new Model Entity in the Models service database.
+ The Model Entity will be registered for use within the platform.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Name of the model entity. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ project : typing.Optional[str]
+ The URN of the project associated with this model entity
+
+ description : typing.Optional[str]
+ Optional description of the model
+
+ spec : typing.Optional[ModelSpec]
+ Detailed specification for the model - Automatically generated by the platform at creation when fileset provided.
+
+ finetuning_type : typing.Optional[FinetuningType]
+ Set for full weight finetuned models
+
+ fileset : typing.Optional[str]
+ A set of checkpoint files, configs, and other auxiliary info associated with this model - expected format {workspace}/{fileset_name}
+
+ base_model : typing.Optional[str]
+ Link to another model which is used as a base for the current model
+
+ api_endpoint : typing.Optional[ApiEndpointData]
+ Data about the inference endpoint for this model
+
+ backend_format : typing.Optional[BackendFormat]
+ Inference API wire format expected by the backend. If unset, inference routing treats the model as OPENAI_CHAT.
+
+ prompt : typing.Optional[PromptData]
+ Configuration for prompt engineering
+
+ custom_fields : typing.Optional[typing.Dict[str, typing.Any]]
+ Custom fields for additional metadata
+
+ ownership : typing.Optional[typing.Dict[str, typing.Any]]
+ Ownership information for the model
+
+ model_providers : typing.Optional[typing.Sequence[str]]
+ List of ModelProvider workspace/name resource names that provide inference for this Model Entity
+
+ trust_remote_code : typing.Optional[bool]
+ Whether to trust remote code for the checkpoint.
+ Some models without support in certain libraries such as Transformers require additional custom Python code to execute.
+ Due to security ramifications of running arbitrary code, this can only be set to true on one of the following conditions:
+ (1) the model's fileset's source is pre-approved in the platform config, or
+ (2) the user creating this model is an administrator.
+
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelEntity
+ Create a new model entity
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.models.create_model(
+ workspace="workspace",
+ name="llama-3.1-8b",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.create_model(
+ workspace,
+ name=name,
+ project=project,
+ description=description,
+ spec=spec,
+ finetuning_type=finetuning_type,
+ fileset=fileset,
+ base_model=base_model,
+ api_endpoint=api_endpoint,
+ backend_format=backend_format,
+ prompt=prompt,
+ custom_fields=custom_fields,
+ ownership=ownership,
+ model_providers=model_providers,
+ trust_remote_code=trust_remote_code,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def create_model_adapter(
+ self,
+ workspace: str,
+ model_name: str,
+ *,
+ name: str,
+ fileset: str,
+ finetuning_type: FinetuningType,
+ description: typing.Optional[str] = OMIT,
+ enabled: typing.Optional[bool] = OMIT,
+ lora_config: typing.Optional[Lora] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Adapter:
+ """
+ Adds an Adapter to the Model
+
+ Parameters
+ ----------
+ workspace : str
+
+ model_name : str
+
+ name : str
+ Name of the adapter. Name must be unique in the workspace. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ fileset : str
+ Location where adapter files are stored - expected format {workspace}/{fileset_name}
+
+ finetuning_type : FinetuningType
+ Type of finetuning (LORA, P_TUNING, etc.)
+
+ description : typing.Optional[str]
+ Optional description of the adapter
+
+ enabled : typing.Optional[bool]
+ Whether to make this adapter available for inference post training
+
+ lora_config : typing.Optional[Lora]
+ Lora configuration specifics
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Adapter
+ Register a new adapter to the model
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.models.create_model_adapter(
+ workspace="workspace",
+ model_name="model_name",
+ name="lora-adapter-v1",
+ fileset="fileset",
+ finetuning_type="lora_merged",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.create_model_adapter(
+ workspace,
+ model_name,
+ name=name,
+ fileset=fileset,
+ finetuning_type=finetuning_type,
+ description=description,
+ enabled=enabled,
+ lora_config=lora_config,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def delete_model_adapter(
+ self, workspace: str, model_name: str, adapter: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> None:
+ """
+ Delete Adapter from Model entity.
+
+ Permanently deletes an adapter from a model entity, if it was deployed, it will be cleaned up automatically.
+
+ Parameters
+ ----------
+ workspace : str
+
+ model_name : str
+
+ adapter : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.models.delete_model_adapter(
+ workspace="workspace",
+ model_name="model_name",
+ adapter="adapter",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.delete_model_adapter(
+ workspace, model_name, adapter, request_options=request_options
+ )
+ return _response.data
+
+ async def update_model_adapter(
+ self,
+ workspace: str,
+ model_name: str,
+ adapter: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ enabled: typing.Optional[bool] = OMIT,
+ fileset: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> Adapter:
+ """
+ Update Adapter deployment or description.
+
+ Parameters
+ ----------
+ workspace : str
+
+ model_name : str
+
+ adapter : str
+
+ description : typing.Optional[str]
+ Optional description of the adapter
+
+ enabled : typing.Optional[bool]
+ Whether to make this adapter available for inference post training
+
+ fileset : typing.Optional[str]
+ Updated fileset for the adapter
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ Adapter
+ Update adapter metadata
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.models.update_model_adapter(
+ workspace="workspace",
+ model_name="model_name",
+ adapter="adapter",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.update_model_adapter(
+ workspace,
+ model_name,
+ adapter,
+ description=description,
+ enabled=enabled,
+ fileset=fileset,
+ request_options=request_options,
+ )
+ return _response.data
+
+ async def get_model(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ verbose: typing.Optional[bool] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelEntity:
+ """
+ Get Model by Workspace and Name.
+
+ Returns the details of a specific model entity identified by its workspace and name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ verbose : typing.Optional[bool]
+ Whether to include full spec details
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelEntity
+ Return model details
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.models.get_model(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.get_model(workspace, name, verbose=verbose, request_options=request_options)
+ return _response.data
+
+ async def delete_model(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> None:
+ """
+ Delete Model entity.
+
+ Permanently deletes a model entity from the platform.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ None
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.models.delete_model(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.delete_model(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def update_model(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ verbose: typing.Optional[bool] = None,
+ description: typing.Optional[str] = OMIT,
+ spec: typing.Optional[ModelSpec] = OMIT,
+ fileset: typing.Optional[str] = OMIT,
+ finetuning_type: typing.Optional[FinetuningType] = OMIT,
+ base_model: typing.Optional[str] = OMIT,
+ api_endpoint: typing.Optional[ApiEndpointData] = OMIT,
+ backend_format: typing.Optional[BackendFormat] = OMIT,
+ prompt: typing.Optional[PromptData] = OMIT,
+ custom_fields: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ ownership: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ model_providers: typing.Optional[typing.Sequence[str]] = OMIT,
+ trust_remote_code: typing.Optional[bool] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> ModelEntity:
+ """
+ Update Model metadata.
+
+ Updates the metadata of an existing model entity. If the request body has an empty field,
+ the old value is kept.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ verbose : typing.Optional[bool]
+ Whether to include full spec details
+
+ description : typing.Optional[str]
+ Optional description of the model
+
+ spec : typing.Optional[ModelSpec]
+ Detailed specification for the model
+
+ fileset : typing.Optional[str]
+ A set of checkpoint files, configs, and other auxiliary info associated with this model - expected format {workspace}/{fileset_name}
+
+ finetuning_type : typing.Optional[FinetuningType]
+ Set for full weight finetuned models
+
+ base_model : typing.Optional[str]
+ Link to another model which is used as a base for the current model
+
+ api_endpoint : typing.Optional[ApiEndpointData]
+ Data about the inference endpoint for this model
+
+ backend_format : typing.Optional[BackendFormat]
+ Inference API wire format expected by the backend. If unset, inference routing treats the model as OPENAI_CHAT.
+
+ prompt : typing.Optional[PromptData]
+ Configuration for prompt engineering
+
+ custom_fields : typing.Optional[typing.Dict[str, typing.Any]]
+ Custom fields for additional metadata
+
+ ownership : typing.Optional[typing.Dict[str, typing.Any]]
+ Ownership information for the model
+
+ model_providers : typing.Optional[typing.Sequence[str]]
+ List of ModelProvider workspace/name resource names that provide inference for this Model Entity
+
+ trust_remote_code : typing.Optional[bool]
+ Whether to trust remote code for the checkpoint.
+ Some models without support in certain libraries such as Transformers require additional custom Python code to execute.
+ Due to security ramifications of running arbitrary code, this can only be set to true on one of the following conditions:
+ (1) the model's fileset's source is pre-approved in the platform config, or
+ (2) the user creating this model is an administrator.
+
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ ModelEntity
+ Update model metadata
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.models.update_model(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.update_model(
+ workspace,
+ name,
+ verbose=verbose,
+ description=description,
+ spec=spec,
+ fileset=fileset,
+ finetuning_type=finetuning_type,
+ base_model=base_model,
+ api_endpoint=api_endpoint,
+ backend_format=backend_format,
+ prompt=prompt,
+ custom_fields=custom_fields,
+ ownership=ownership,
+ model_providers=model_providers,
+ trust_remote_code=trust_remote_code,
+ request_options=request_options,
+ )
+ return _response.data
diff --git a/sdks/python/models/raw_client.py b/sdks/python/models/raw_client.py
new file mode 100644
index 0000000000..c8fe9e6f98
--- /dev/null
+++ b/sdks/python/models/raw_client.py
@@ -0,0 +1,1487 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+from json.decoder import JSONDecodeError
+
+from ..core.api_error import ApiError
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.http_response import AsyncHttpResponse, HttpResponse
+from ..core.jsonable_encoder import encode_path_param
+from ..core.parse_error import ParsingError
+from ..core.pydantic_utilities import parse_obj_as
+from ..core.request_options import RequestOptions
+from ..core.serialization import convert_and_respect_annotation_metadata
+from ..errors.unprocessable_entity_error import UnprocessableEntityError
+from ..types.adapter import Adapter
+from ..types.api_endpoint_data import ApiEndpointData
+from ..types.backend_format import BackendFormat
+from ..types.finetuning_type import FinetuningType
+from ..types.lora import Lora
+from ..types.model_entity import ModelEntity
+from ..types.model_entity_filter import ModelEntityFilter
+from ..types.model_entity_sort_field import ModelEntitySortField
+from ..types.model_entitys_page import ModelEntitysPage
+from ..types.model_spec import ModelSpec
+from ..types.prompt_data import PromptData
+from pydantic import ValidationError
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class RawModelsClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ def list_models(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[ModelEntitySortField] = None,
+ verbose: typing.Optional[bool] = None,
+ filter: typing.Optional[ModelEntityFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ModelEntitysPage]:
+ """
+ List Models endpoint with filtering, pagination, and sorting.
+
+ Supports filter parameters for various criteria (including peft, custom fields),
+ pagination (page, page_size), sorting, and workspace filtering via query parameter.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[ModelEntitySortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ verbose : typing.Optional[bool]
+ Whether to include full spec details
+
+ filter : typing.Optional[ModelEntityFilter]
+ Filter models by name, project, workspace, base_model, adapters, finetuning_type, prompt, lora_enabled, description, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ModelEntitysPage]
+ Return a list of models
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/models",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "verbose": verbose,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=ModelEntityFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelEntitysPage,
+ parse_obj_as(
+ type_=ModelEntitysPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def create_model(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ project: typing.Optional[str] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ spec: typing.Optional[ModelSpec] = OMIT,
+ finetuning_type: typing.Optional[FinetuningType] = OMIT,
+ fileset: typing.Optional[str] = OMIT,
+ base_model: typing.Optional[str] = OMIT,
+ api_endpoint: typing.Optional[ApiEndpointData] = OMIT,
+ backend_format: typing.Optional[BackendFormat] = OMIT,
+ prompt: typing.Optional[PromptData] = OMIT,
+ custom_fields: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ ownership: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ model_providers: typing.Optional[typing.Sequence[str]] = OMIT,
+ trust_remote_code: typing.Optional[bool] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ModelEntity]:
+ """
+ Create a new model entity.
+
+ This endpoint creates a new Model Entity in the Models service database.
+ The Model Entity will be registered for use within the platform.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Name of the model entity. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ project : typing.Optional[str]
+ The URN of the project associated with this model entity
+
+ description : typing.Optional[str]
+ Optional description of the model
+
+ spec : typing.Optional[ModelSpec]
+ Detailed specification for the model - Automatically generated by the platform at creation when fileset provided.
+
+ finetuning_type : typing.Optional[FinetuningType]
+ Set for full weight finetuned models
+
+ fileset : typing.Optional[str]
+ A set of checkpoint files, configs, and other auxiliary info associated with this model - expected format {workspace}/{fileset_name}
+
+ base_model : typing.Optional[str]
+ Link to another model which is used as a base for the current model
+
+ api_endpoint : typing.Optional[ApiEndpointData]
+ Data about the inference endpoint for this model
+
+ backend_format : typing.Optional[BackendFormat]
+ Inference API wire format expected by the backend. If unset, inference routing treats the model as OPENAI_CHAT.
+
+ prompt : typing.Optional[PromptData]
+ Configuration for prompt engineering
+
+ custom_fields : typing.Optional[typing.Dict[str, typing.Any]]
+ Custom fields for additional metadata
+
+ ownership : typing.Optional[typing.Dict[str, typing.Any]]
+ Ownership information for the model
+
+ model_providers : typing.Optional[typing.Sequence[str]]
+ List of ModelProvider workspace/name resource names that provide inference for this Model Entity
+
+ trust_remote_code : typing.Optional[bool]
+ Whether to trust remote code for the checkpoint.
+ Some models without support in certain libraries such as Transformers require additional custom Python code to execute.
+ Due to security ramifications of running arbitrary code, this can only be set to true on one of the following conditions:
+ (1) the model's fileset's source is pre-approved in the platform config, or
+ (2) the user creating this model is an administrator.
+
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ModelEntity]
+ Create a new model entity
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/models",
+ method="POST",
+ json={
+ "name": name,
+ "project": project,
+ "description": description,
+ "spec": convert_and_respect_annotation_metadata(object_=spec, annotation=ModelSpec, direction="write"),
+ "finetuning_type": finetuning_type,
+ "fileset": fileset,
+ "base_model": base_model,
+ "api_endpoint": convert_and_respect_annotation_metadata(
+ object_=api_endpoint, annotation=ApiEndpointData, direction="write"
+ ),
+ "backend_format": backend_format,
+ "prompt": convert_and_respect_annotation_metadata(
+ object_=prompt, annotation=PromptData, direction="write"
+ ),
+ "custom_fields": custom_fields,
+ "ownership": ownership,
+ "model_providers": model_providers,
+ "trust_remote_code": trust_remote_code,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelEntity,
+ parse_obj_as(
+ type_=ModelEntity, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def create_model_adapter(
+ self,
+ workspace: str,
+ model_name: str,
+ *,
+ name: str,
+ fileset: str,
+ finetuning_type: FinetuningType,
+ description: typing.Optional[str] = OMIT,
+ enabled: typing.Optional[bool] = OMIT,
+ lora_config: typing.Optional[Lora] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[Adapter]:
+ """
+ Adds an Adapter to the Model
+
+ Parameters
+ ----------
+ workspace : str
+
+ model_name : str
+
+ name : str
+ Name of the adapter. Name must be unique in the workspace. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ fileset : str
+ Location where adapter files are stored - expected format {workspace}/{fileset_name}
+
+ finetuning_type : FinetuningType
+ Type of finetuning (LORA, P_TUNING, etc.)
+
+ description : typing.Optional[str]
+ Optional description of the adapter
+
+ enabled : typing.Optional[bool]
+ Whether to make this adapter available for inference post training
+
+ lora_config : typing.Optional[Lora]
+ Lora configuration specifics
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[Adapter]
+ Register a new adapter to the model
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/models/{encode_path_param(model_name)}/adapters",
+ method="POST",
+ json={
+ "name": name,
+ "description": description,
+ "fileset": fileset,
+ "finetuning_type": finetuning_type,
+ "enabled": enabled,
+ "lora_config": convert_and_respect_annotation_metadata(
+ object_=lora_config, annotation=Lora, direction="write"
+ ),
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Adapter,
+ parse_obj_as(
+ type_=Adapter, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def delete_model_adapter(
+ self, workspace: str, model_name: str, adapter: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[None]:
+ """
+ Delete Adapter from Model entity.
+
+ Permanently deletes an adapter from a model entity, if it was deployed, it will be cleaned up automatically.
+
+ Parameters
+ ----------
+ workspace : str
+
+ model_name : str
+
+ adapter : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[None]
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/models/{encode_path_param(model_name)}/adapters/{encode_path_param(adapter)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return HttpResponse(response=_response, data=None)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def update_model_adapter(
+ self,
+ workspace: str,
+ model_name: str,
+ adapter: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ enabled: typing.Optional[bool] = OMIT,
+ fileset: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[Adapter]:
+ """
+ Update Adapter deployment or description.
+
+ Parameters
+ ----------
+ workspace : str
+
+ model_name : str
+
+ adapter : str
+
+ description : typing.Optional[str]
+ Optional description of the adapter
+
+ enabled : typing.Optional[bool]
+ Whether to make this adapter available for inference post training
+
+ fileset : typing.Optional[str]
+ Updated fileset for the adapter
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[Adapter]
+ Update adapter metadata
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/models/{encode_path_param(model_name)}/adapters/{encode_path_param(adapter)}",
+ method="PATCH",
+ json={
+ "description": description,
+ "enabled": enabled,
+ "fileset": fileset,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Adapter,
+ parse_obj_as(
+ type_=Adapter, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def get_model(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ verbose: typing.Optional[bool] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ModelEntity]:
+ """
+ Get Model by Workspace and Name.
+
+ Returns the details of a specific model entity identified by its workspace and name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ verbose : typing.Optional[bool]
+ Whether to include full spec details
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ModelEntity]
+ Return model details
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/models/{encode_path_param(name)}",
+ method="GET",
+ params={
+ "verbose": verbose,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelEntity,
+ parse_obj_as(
+ type_=ModelEntity, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def delete_model(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[None]:
+ """
+ Delete Model entity.
+
+ Permanently deletes a model entity from the platform.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[None]
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/models/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return HttpResponse(response=_response, data=None)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def update_model(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ verbose: typing.Optional[bool] = None,
+ description: typing.Optional[str] = OMIT,
+ spec: typing.Optional[ModelSpec] = OMIT,
+ fileset: typing.Optional[str] = OMIT,
+ finetuning_type: typing.Optional[FinetuningType] = OMIT,
+ base_model: typing.Optional[str] = OMIT,
+ api_endpoint: typing.Optional[ApiEndpointData] = OMIT,
+ backend_format: typing.Optional[BackendFormat] = OMIT,
+ prompt: typing.Optional[PromptData] = OMIT,
+ custom_fields: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ ownership: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ model_providers: typing.Optional[typing.Sequence[str]] = OMIT,
+ trust_remote_code: typing.Optional[bool] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[ModelEntity]:
+ """
+ Update Model metadata.
+
+ Updates the metadata of an existing model entity. If the request body has an empty field,
+ the old value is kept.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ verbose : typing.Optional[bool]
+ Whether to include full spec details
+
+ description : typing.Optional[str]
+ Optional description of the model
+
+ spec : typing.Optional[ModelSpec]
+ Detailed specification for the model
+
+ fileset : typing.Optional[str]
+ A set of checkpoint files, configs, and other auxiliary info associated with this model - expected format {workspace}/{fileset_name}
+
+ finetuning_type : typing.Optional[FinetuningType]
+ Set for full weight finetuned models
+
+ base_model : typing.Optional[str]
+ Link to another model which is used as a base for the current model
+
+ api_endpoint : typing.Optional[ApiEndpointData]
+ Data about the inference endpoint for this model
+
+ backend_format : typing.Optional[BackendFormat]
+ Inference API wire format expected by the backend. If unset, inference routing treats the model as OPENAI_CHAT.
+
+ prompt : typing.Optional[PromptData]
+ Configuration for prompt engineering
+
+ custom_fields : typing.Optional[typing.Dict[str, typing.Any]]
+ Custom fields for additional metadata
+
+ ownership : typing.Optional[typing.Dict[str, typing.Any]]
+ Ownership information for the model
+
+ model_providers : typing.Optional[typing.Sequence[str]]
+ List of ModelProvider workspace/name resource names that provide inference for this Model Entity
+
+ trust_remote_code : typing.Optional[bool]
+ Whether to trust remote code for the checkpoint.
+ Some models without support in certain libraries such as Transformers require additional custom Python code to execute.
+ Due to security ramifications of running arbitrary code, this can only be set to true on one of the following conditions:
+ (1) the model's fileset's source is pre-approved in the platform config, or
+ (2) the user creating this model is an administrator.
+
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[ModelEntity]
+ Update model metadata
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/models/{encode_path_param(name)}",
+ method="PATCH",
+ params={
+ "verbose": verbose,
+ },
+ json={
+ "description": description,
+ "spec": convert_and_respect_annotation_metadata(object_=spec, annotation=ModelSpec, direction="write"),
+ "fileset": fileset,
+ "finetuning_type": finetuning_type,
+ "base_model": base_model,
+ "api_endpoint": convert_and_respect_annotation_metadata(
+ object_=api_endpoint, annotation=ApiEndpointData, direction="write"
+ ),
+ "backend_format": backend_format,
+ "prompt": convert_and_respect_annotation_metadata(
+ object_=prompt, annotation=PromptData, direction="write"
+ ),
+ "custom_fields": custom_fields,
+ "ownership": ownership,
+ "model_providers": model_providers,
+ "trust_remote_code": trust_remote_code,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelEntity,
+ parse_obj_as(
+ type_=ModelEntity, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+
+class AsyncRawModelsClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ async def list_models(
+ self,
+ workspace: str,
+ *,
+ page: typing.Optional[int] = None,
+ page_size: typing.Optional[int] = None,
+ sort: typing.Optional[ModelEntitySortField] = None,
+ verbose: typing.Optional[bool] = None,
+ filter: typing.Optional[ModelEntityFilter] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ModelEntitysPage]:
+ """
+ List Models endpoint with filtering, pagination, and sorting.
+
+ Supports filter parameters for various criteria (including peft, custom fields),
+ pagination (page, page_size), sorting, and workspace filtering via query parameter.
+
+ Parameters
+ ----------
+ workspace : str
+
+ page : typing.Optional[int]
+ Page number.
+
+ page_size : typing.Optional[int]
+ Page size.
+
+ sort : typing.Optional[ModelEntitySortField]
+ The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+ verbose : typing.Optional[bool]
+ Whether to include full spec details
+
+ filter : typing.Optional[ModelEntityFilter]
+ Filter models by name, project, workspace, base_model, adapters, finetuning_type, prompt, lora_enabled, description, created_at, and updated_at.
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ModelEntitysPage]
+ Return a list of models
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/models",
+ method="GET",
+ params={
+ "page": page,
+ "page_size": page_size,
+ "sort": sort,
+ "verbose": verbose,
+ "filter": convert_and_respect_annotation_metadata(
+ object_=filter, annotation=ModelEntityFilter, direction="write"
+ ),
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelEntitysPage,
+ parse_obj_as(
+ type_=ModelEntitysPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def create_model(
+ self,
+ workspace: str,
+ *,
+ name: str,
+ project: typing.Optional[str] = OMIT,
+ description: typing.Optional[str] = OMIT,
+ spec: typing.Optional[ModelSpec] = OMIT,
+ finetuning_type: typing.Optional[FinetuningType] = OMIT,
+ fileset: typing.Optional[str] = OMIT,
+ base_model: typing.Optional[str] = OMIT,
+ api_endpoint: typing.Optional[ApiEndpointData] = OMIT,
+ backend_format: typing.Optional[BackendFormat] = OMIT,
+ prompt: typing.Optional[PromptData] = OMIT,
+ custom_fields: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ ownership: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ model_providers: typing.Optional[typing.Sequence[str]] = OMIT,
+ trust_remote_code: typing.Optional[bool] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ModelEntity]:
+ """
+ Create a new model entity.
+
+ This endpoint creates a new Model Entity in the Models service database.
+ The Model Entity will be registered for use within the platform.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+ Name of the model entity. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ project : typing.Optional[str]
+ The URN of the project associated with this model entity
+
+ description : typing.Optional[str]
+ Optional description of the model
+
+ spec : typing.Optional[ModelSpec]
+ Detailed specification for the model - Automatically generated by the platform at creation when fileset provided.
+
+ finetuning_type : typing.Optional[FinetuningType]
+ Set for full weight finetuned models
+
+ fileset : typing.Optional[str]
+ A set of checkpoint files, configs, and other auxiliary info associated with this model - expected format {workspace}/{fileset_name}
+
+ base_model : typing.Optional[str]
+ Link to another model which is used as a base for the current model
+
+ api_endpoint : typing.Optional[ApiEndpointData]
+ Data about the inference endpoint for this model
+
+ backend_format : typing.Optional[BackendFormat]
+ Inference API wire format expected by the backend. If unset, inference routing treats the model as OPENAI_CHAT.
+
+ prompt : typing.Optional[PromptData]
+ Configuration for prompt engineering
+
+ custom_fields : typing.Optional[typing.Dict[str, typing.Any]]
+ Custom fields for additional metadata
+
+ ownership : typing.Optional[typing.Dict[str, typing.Any]]
+ Ownership information for the model
+
+ model_providers : typing.Optional[typing.Sequence[str]]
+ List of ModelProvider workspace/name resource names that provide inference for this Model Entity
+
+ trust_remote_code : typing.Optional[bool]
+ Whether to trust remote code for the checkpoint.
+ Some models without support in certain libraries such as Transformers require additional custom Python code to execute.
+ Due to security ramifications of running arbitrary code, this can only be set to true on one of the following conditions:
+ (1) the model's fileset's source is pre-approved in the platform config, or
+ (2) the user creating this model is an administrator.
+
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ModelEntity]
+ Create a new model entity
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/models",
+ method="POST",
+ json={
+ "name": name,
+ "project": project,
+ "description": description,
+ "spec": convert_and_respect_annotation_metadata(object_=spec, annotation=ModelSpec, direction="write"),
+ "finetuning_type": finetuning_type,
+ "fileset": fileset,
+ "base_model": base_model,
+ "api_endpoint": convert_and_respect_annotation_metadata(
+ object_=api_endpoint, annotation=ApiEndpointData, direction="write"
+ ),
+ "backend_format": backend_format,
+ "prompt": convert_and_respect_annotation_metadata(
+ object_=prompt, annotation=PromptData, direction="write"
+ ),
+ "custom_fields": custom_fields,
+ "ownership": ownership,
+ "model_providers": model_providers,
+ "trust_remote_code": trust_remote_code,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelEntity,
+ parse_obj_as(
+ type_=ModelEntity, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def create_model_adapter(
+ self,
+ workspace: str,
+ model_name: str,
+ *,
+ name: str,
+ fileset: str,
+ finetuning_type: FinetuningType,
+ description: typing.Optional[str] = OMIT,
+ enabled: typing.Optional[bool] = OMIT,
+ lora_config: typing.Optional[Lora] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[Adapter]:
+ """
+ Adds an Adapter to the Model
+
+ Parameters
+ ----------
+ workspace : str
+
+ model_name : str
+
+ name : str
+ Name of the adapter. Name must be unique in the workspace. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+ fileset : str
+ Location where adapter files are stored - expected format {workspace}/{fileset_name}
+
+ finetuning_type : FinetuningType
+ Type of finetuning (LORA, P_TUNING, etc.)
+
+ description : typing.Optional[str]
+ Optional description of the adapter
+
+ enabled : typing.Optional[bool]
+ Whether to make this adapter available for inference post training
+
+ lora_config : typing.Optional[Lora]
+ Lora configuration specifics
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[Adapter]
+ Register a new adapter to the model
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/models/{encode_path_param(model_name)}/adapters",
+ method="POST",
+ json={
+ "name": name,
+ "description": description,
+ "fileset": fileset,
+ "finetuning_type": finetuning_type,
+ "enabled": enabled,
+ "lora_config": convert_and_respect_annotation_metadata(
+ object_=lora_config, annotation=Lora, direction="write"
+ ),
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Adapter,
+ parse_obj_as(
+ type_=Adapter, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def delete_model_adapter(
+ self, workspace: str, model_name: str, adapter: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[None]:
+ """
+ Delete Adapter from Model entity.
+
+ Permanently deletes an adapter from a model entity, if it was deployed, it will be cleaned up automatically.
+
+ Parameters
+ ----------
+ workspace : str
+
+ model_name : str
+
+ adapter : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[None]
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/models/{encode_path_param(model_name)}/adapters/{encode_path_param(adapter)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return AsyncHttpResponse(response=_response, data=None)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def update_model_adapter(
+ self,
+ workspace: str,
+ model_name: str,
+ adapter: str,
+ *,
+ description: typing.Optional[str] = OMIT,
+ enabled: typing.Optional[bool] = OMIT,
+ fileset: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[Adapter]:
+ """
+ Update Adapter deployment or description.
+
+ Parameters
+ ----------
+ workspace : str
+
+ model_name : str
+
+ adapter : str
+
+ description : typing.Optional[str]
+ Optional description of the adapter
+
+ enabled : typing.Optional[bool]
+ Whether to make this adapter available for inference post training
+
+ fileset : typing.Optional[str]
+ Updated fileset for the adapter
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[Adapter]
+ Update adapter metadata
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/models/{encode_path_param(model_name)}/adapters/{encode_path_param(adapter)}",
+ method="PATCH",
+ json={
+ "description": description,
+ "enabled": enabled,
+ "fileset": fileset,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ Adapter,
+ parse_obj_as(
+ type_=Adapter, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def get_model(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ verbose: typing.Optional[bool] = None,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ModelEntity]:
+ """
+ Get Model by Workspace and Name.
+
+ Returns the details of a specific model entity identified by its workspace and name.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ verbose : typing.Optional[bool]
+ Whether to include full spec details
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ModelEntity]
+ Return model details
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/models/{encode_path_param(name)}",
+ method="GET",
+ params={
+ "verbose": verbose,
+ },
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelEntity,
+ parse_obj_as(
+ type_=ModelEntity, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def delete_model(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[None]:
+ """
+ Delete Model entity.
+
+ Permanently deletes a model entity from the platform.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[None]
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/models/{encode_path_param(name)}",
+ method="DELETE",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ return AsyncHttpResponse(response=_response, data=None)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def update_model(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ verbose: typing.Optional[bool] = None,
+ description: typing.Optional[str] = OMIT,
+ spec: typing.Optional[ModelSpec] = OMIT,
+ fileset: typing.Optional[str] = OMIT,
+ finetuning_type: typing.Optional[FinetuningType] = OMIT,
+ base_model: typing.Optional[str] = OMIT,
+ api_endpoint: typing.Optional[ApiEndpointData] = OMIT,
+ backend_format: typing.Optional[BackendFormat] = OMIT,
+ prompt: typing.Optional[PromptData] = OMIT,
+ custom_fields: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ ownership: typing.Optional[typing.Dict[str, typing.Any]] = OMIT,
+ model_providers: typing.Optional[typing.Sequence[str]] = OMIT,
+ trust_remote_code: typing.Optional[bool] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[ModelEntity]:
+ """
+ Update Model metadata.
+
+ Updates the metadata of an existing model entity. If the request body has an empty field,
+ the old value is kept.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ verbose : typing.Optional[bool]
+ Whether to include full spec details
+
+ description : typing.Optional[str]
+ Optional description of the model
+
+ spec : typing.Optional[ModelSpec]
+ Detailed specification for the model
+
+ fileset : typing.Optional[str]
+ A set of checkpoint files, configs, and other auxiliary info associated with this model - expected format {workspace}/{fileset_name}
+
+ finetuning_type : typing.Optional[FinetuningType]
+ Set for full weight finetuned models
+
+ base_model : typing.Optional[str]
+ Link to another model which is used as a base for the current model
+
+ api_endpoint : typing.Optional[ApiEndpointData]
+ Data about the inference endpoint for this model
+
+ backend_format : typing.Optional[BackendFormat]
+ Inference API wire format expected by the backend. If unset, inference routing treats the model as OPENAI_CHAT.
+
+ prompt : typing.Optional[PromptData]
+ Configuration for prompt engineering
+
+ custom_fields : typing.Optional[typing.Dict[str, typing.Any]]
+ Custom fields for additional metadata
+
+ ownership : typing.Optional[typing.Dict[str, typing.Any]]
+ Ownership information for the model
+
+ model_providers : typing.Optional[typing.Sequence[str]]
+ List of ModelProvider workspace/name resource names that provide inference for this Model Entity
+
+ trust_remote_code : typing.Optional[bool]
+ Whether to trust remote code for the checkpoint.
+ Some models without support in certain libraries such as Transformers require additional custom Python code to execute.
+ Due to security ramifications of running arbitrary code, this can only be set to true on one of the following conditions:
+ (1) the model's fileset's source is pre-approved in the platform config, or
+ (2) the user creating this model is an administrator.
+
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[ModelEntity]
+ Update model metadata
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/models/v2/workspaces/{encode_path_param(workspace)}/models/{encode_path_param(name)}",
+ method="PATCH",
+ params={
+ "verbose": verbose,
+ },
+ json={
+ "description": description,
+ "spec": convert_and_respect_annotation_metadata(object_=spec, annotation=ModelSpec, direction="write"),
+ "fileset": fileset,
+ "finetuning_type": finetuning_type,
+ "base_model": base_model,
+ "api_endpoint": convert_and_respect_annotation_metadata(
+ object_=api_endpoint, annotation=ApiEndpointData, direction="write"
+ ),
+ "backend_format": backend_format,
+ "prompt": convert_and_respect_annotation_metadata(
+ object_=prompt, annotation=PromptData, direction="write"
+ ),
+ "custom_fields": custom_fields,
+ "ownership": ownership,
+ "model_providers": model_providers,
+ "trust_remote_code": trust_remote_code,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ ModelEntity,
+ parse_obj_as(
+ type_=ModelEntity, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
diff --git a/sdks/python/otlp/__init__.py b/sdks/python/otlp/__init__.py
new file mode 100644
index 0000000000..5cde0202dc
--- /dev/null
+++ b/sdks/python/otlp/__init__.py
@@ -0,0 +1,4 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+
diff --git a/sdks/python/otlp/client.py b/sdks/python/otlp/client.py
new file mode 100644
index 0000000000..2c43d65678
--- /dev/null
+++ b/sdks/python/otlp/client.py
@@ -0,0 +1,246 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.request_options import RequestOptions
+from ..types.otel_export_logs_service_response import OtelExportLogsServiceResponse
+from ..types.platform_job_log_page import PlatformJobLogPage
+from .raw_client import AsyncRawOtlpClient, RawOtlpClient
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class OtlpClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._raw_client = RawOtlpClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> RawOtlpClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ RawOtlpClient
+ """
+ return self._raw_client
+
+ def upload_otlp_logs(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> OtelExportLogsServiceResponse:
+ """
+ Upload OTLP logs to a specified fileset in JSON or Protobuf format.
+
+ Supports both application/json and application/x-protobuf content types.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ OtelExportLogsServiceResponse
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.otlp.upload_otlp_logs(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.upload_otlp_logs(workspace, name, request_options=request_options)
+ return _response.data
+
+ def query_otlp_logs(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ filters: typing.Optional[typing.Dict[str, str]] = OMIT,
+ limit: typing.Optional[int] = OMIT,
+ page_cursor: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> PlatformJobLogPage:
+ """
+ Query logs from parquet files in a fileset.
+
+ This is an internal endpoint that runs DuckDB queries with direct storage
+ access.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ filters : typing.Optional[typing.Dict[str, str]]
+ Key-value filters to apply to the query
+
+ limit : typing.Optional[int]
+ Maximum number of results to return
+
+ page_cursor : typing.Optional[str]
+ Cursor for pagination
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobLogPage
+ Successful Response
+
+ Examples
+ --------
+ from nvidia import NvidiaApi
+
+ client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+ client.otlp.query_otlp_logs(
+ workspace="workspace",
+ name="name",
+ )
+ """
+ _response = self._raw_client.query_otlp_logs(
+ workspace, name, filters=filters, limit=limit, page_cursor=page_cursor, request_options=request_options
+ )
+ return _response.data
+
+
+class AsyncOtlpClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._raw_client = AsyncRawOtlpClient(client_wrapper=client_wrapper)
+
+ @property
+ def with_raw_response(self) -> AsyncRawOtlpClient:
+ """
+ Retrieves a raw implementation of this client that returns raw responses.
+
+ Returns
+ -------
+ AsyncRawOtlpClient
+ """
+ return self._raw_client
+
+ async def upload_otlp_logs(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> OtelExportLogsServiceResponse:
+ """
+ Upload OTLP logs to a specified fileset in JSON or Protobuf format.
+
+ Supports both application/json and application/x-protobuf content types.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ OtelExportLogsServiceResponse
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.otlp.upload_otlp_logs(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.upload_otlp_logs(workspace, name, request_options=request_options)
+ return _response.data
+
+ async def query_otlp_logs(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ filters: typing.Optional[typing.Dict[str, str]] = OMIT,
+ limit: typing.Optional[int] = OMIT,
+ page_cursor: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> PlatformJobLogPage:
+ """
+ Query logs from parquet files in a fileset.
+
+ This is an internal endpoint that runs DuckDB queries with direct storage
+ access.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ filters : typing.Optional[typing.Dict[str, str]]
+ Key-value filters to apply to the query
+
+ limit : typing.Optional[int]
+ Maximum number of results to return
+
+ page_cursor : typing.Optional[str]
+ Cursor for pagination
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ PlatformJobLogPage
+ Successful Response
+
+ Examples
+ --------
+ import asyncio
+
+ from nvidia import AsyncNvidiaApi
+
+ client = AsyncNvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+ )
+
+
+ async def main() -> None:
+ await client.otlp.query_otlp_logs(
+ workspace="workspace",
+ name="name",
+ )
+
+
+ asyncio.run(main())
+ """
+ _response = await self._raw_client.query_otlp_logs(
+ workspace, name, filters=filters, limit=limit, page_cursor=page_cursor, request_options=request_options
+ )
+ return _response.data
diff --git a/sdks/python/otlp/raw_client.py b/sdks/python/otlp/raw_client.py
new file mode 100644
index 0000000000..3f717ad9da
--- /dev/null
+++ b/sdks/python/otlp/raw_client.py
@@ -0,0 +1,309 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+from json.decoder import JSONDecodeError
+
+from ..core.api_error import ApiError
+from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
+from ..core.http_response import AsyncHttpResponse, HttpResponse
+from ..core.jsonable_encoder import encode_path_param
+from ..core.parse_error import ParsingError
+from ..core.pydantic_utilities import parse_obj_as
+from ..core.request_options import RequestOptions
+from ..errors.unprocessable_entity_error import UnprocessableEntityError
+from ..types.otel_export_logs_service_response import OtelExportLogsServiceResponse
+from ..types.platform_job_log_page import PlatformJobLogPage
+from pydantic import ValidationError
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class RawOtlpClient:
+ def __init__(self, *, client_wrapper: SyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ def upload_otlp_logs(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> HttpResponse[OtelExportLogsServiceResponse]:
+ """
+ Upload OTLP logs to a specified fileset in JSON or Protobuf format.
+
+ Supports both application/json and application/x-protobuf content types.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[OtelExportLogsServiceResponse]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets/{encode_path_param(name)}/otlp/v1/logs",
+ method="POST",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ OtelExportLogsServiceResponse,
+ parse_obj_as(
+ type_=OtelExportLogsServiceResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ def query_otlp_logs(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ filters: typing.Optional[typing.Dict[str, str]] = OMIT,
+ limit: typing.Optional[int] = OMIT,
+ page_cursor: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> HttpResponse[PlatformJobLogPage]:
+ """
+ Query logs from parquet files in a fileset.
+
+ This is an internal endpoint that runs DuckDB queries with direct storage
+ access.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ filters : typing.Optional[typing.Dict[str, str]]
+ Key-value filters to apply to the query
+
+ limit : typing.Optional[int]
+ Maximum number of results to return
+
+ page_cursor : typing.Optional[str]
+ Cursor for pagination
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ HttpResponse[PlatformJobLogPage]
+ Successful Response
+ """
+ _response = self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets/{encode_path_param(name)}/otlp/v1/logs/query",
+ method="POST",
+ json={
+ "filters": filters,
+ "limit": limit,
+ "page_cursor": page_cursor,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobLogPage,
+ parse_obj_as(
+ type_=PlatformJobLogPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return HttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+
+class AsyncRawOtlpClient:
+ def __init__(self, *, client_wrapper: AsyncClientWrapper):
+ self._client_wrapper = client_wrapper
+
+ async def upload_otlp_logs(
+ self, workspace: str, name: str, *, request_options: typing.Optional[RequestOptions] = None
+ ) -> AsyncHttpResponse[OtelExportLogsServiceResponse]:
+ """
+ Upload OTLP logs to a specified fileset in JSON or Protobuf format.
+
+ Supports both application/json and application/x-protobuf content types.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[OtelExportLogsServiceResponse]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets/{encode_path_param(name)}/otlp/v1/logs",
+ method="POST",
+ request_options=request_options,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ OtelExportLogsServiceResponse,
+ parse_obj_as(
+ type_=OtelExportLogsServiceResponse, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
+
+ async def query_otlp_logs(
+ self,
+ workspace: str,
+ name: str,
+ *,
+ filters: typing.Optional[typing.Dict[str, str]] = OMIT,
+ limit: typing.Optional[int] = OMIT,
+ page_cursor: typing.Optional[str] = OMIT,
+ request_options: typing.Optional[RequestOptions] = None,
+ ) -> AsyncHttpResponse[PlatformJobLogPage]:
+ """
+ Query logs from parquet files in a fileset.
+
+ This is an internal endpoint that runs DuckDB queries with direct storage
+ access.
+
+ Parameters
+ ----------
+ workspace : str
+
+ name : str
+
+ filters : typing.Optional[typing.Dict[str, str]]
+ Key-value filters to apply to the query
+
+ limit : typing.Optional[int]
+ Maximum number of results to return
+
+ page_cursor : typing.Optional[str]
+ Cursor for pagination
+
+ request_options : typing.Optional[RequestOptions]
+ Request-specific configuration.
+
+ Returns
+ -------
+ AsyncHttpResponse[PlatformJobLogPage]
+ Successful Response
+ """
+ _response = await self._client_wrapper.httpx_client.request(
+ f"apis/files/v2/workspaces/{encode_path_param(workspace)}/filesets/{encode_path_param(name)}/otlp/v1/logs/query",
+ method="POST",
+ json={
+ "filters": filters,
+ "limit": limit,
+ "page_cursor": page_cursor,
+ },
+ headers={
+ "content-type": "application/json",
+ },
+ request_options=request_options,
+ omit=OMIT,
+ )
+ try:
+ if 200 <= _response.status_code < 300:
+ _data = typing.cast(
+ PlatformJobLogPage,
+ parse_obj_as(
+ type_=PlatformJobLogPage, # type: ignore
+ object_=_response.json(),
+ ),
+ )
+ return AsyncHttpResponse(response=_response, data=_data)
+ if _response.status_code == 422:
+ raise UnprocessableEntityError(
+ headers=dict(_response.headers),
+ body=typing.cast(
+ typing.Any,
+ parse_obj_as(
+ type_=typing.Any, # type: ignore
+ object_=_response.json(),
+ ),
+ ),
+ )
+ _response_json = _response.json()
+ except JSONDecodeError:
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response.text)
+ except ValidationError as e:
+ raise ParsingError(
+ status_code=_response.status_code, headers=dict(_response.headers), body=_response.json(), cause=e
+ )
+ raise ApiError(status_code=_response.status_code, headers=dict(_response.headers), body=_response_json)
diff --git a/sdks/python/reference.md b/sdks/python/reference.md
new file mode 100644
index 0000000000..e28b8f9655
--- /dev/null
+++ b/sdks/python/reference.md
@@ -0,0 +1,15286 @@
+# Reference
+## Discovery
+client.discovery.get_auth_discovery() -> AuthDiscoveryResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Return authentication configuration for CLI/SDK discovery.
+
+This endpoint is unauthenticated and returns the information clients
+need to authenticate with this NeMo Platform deployment.
+
+**Response fields:**
+
+- `auth_enabled`: Whether authentication is enabled on this cluster
+- `oidc`: OIDC configuration (only present when OIDC is enabled)
+ - `issuer`: The OIDC issuer URL
+ - `authorization_endpoint`: Authorization endpoint for browser-based flows
+ - `token_endpoint`: Token exchange endpoint
+ - `device_authorization_endpoint`: Device flow authorization endpoint (for CLI)
+ - `userinfo_endpoint`: UserInfo endpoint
+ - `client_id`: OAuth client ID to use
+ - `default_scopes`: OAuth scopes to request during authentication
+ - `scope_prefix`: Prefix to prepend to custom scopes (those with ':' or '.default')
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.discovery.get_auth_discovery()
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.iam.list_role_bindings(...) -> RoleBindingsPage
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+List all role bindings (Platform Admin only)
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.iam.list_role_bindings()
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number.
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Page size.
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[str]` — The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+
+
+
+
+-
+
+**filter:** `typing.Optional[RoleBindingFilter]` — Filter role bindings by principal, workspace, role, granted_by, is_active, granted_at, and revoked_at.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.iam.create_role_binding(...) -> RoleBinding
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Create a new role binding (Platform Admin only)
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.iam.create_role_binding(
+ principal="principal",
+ role="role",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**principal:** `str` — The principal identifier (email, user ID, or group ID)
+
+
+
+
+
+-
+
+**role:** `str` — The role name (e.g., 'Viewer', 'Editor', 'Admin')
+
+
+
+
+
+-
+
+**wait_role_propagation:** `typing.Optional[bool]` — If true, wait for role to propagate before returning (default: true). Set to false for bulk operations.
+
+
+
+
+
+-
+
+**workspace:** `typing.Optional[str]` — The workspace this binding applies to. None for platform-level roles.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.iam.get_role_binding(...) -> RoleBinding
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get a specific role binding (Platform Admin only)
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.iam.get_role_binding(
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.iam.revoke_role_binding(...) -> DeleteResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Revoke a role binding (Platform Admin only)
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.iam.revoke_role_binding(
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**wait_role_propagation:** `typing.Optional[bool]` — If true, wait for role to propagate before returning (default: true). Set to false for bulk operations.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.entity_store.get_entity_by_id(...) -> Entity
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get a specific entity by its unique identifier.
+This endpoint is primarily for debugging and internal use.
+
+Example:
+```
+GET /apis/entities/v2/entities/customization-config-5Q2LoF8z8M9JZxZsHwJKNn
+```
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.entity_store.get_entity_by_id(
+ id="id",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**id:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.entity_store.list_workspaces(...) -> WorkspacesPage
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+List all workspaces with pagination.
+
+When authentication is enabled, only workspaces the principal has access to
+are returned. Service principals and platform admins have access to all workspaces.
+
+Query Parameters:
+- page, page_size: Pagination
+- sort: Sort field
+- filter: Advanced filters (JSON, text, or bracket notation)
+
+Example:
+```
+GET /apis/entities/v2/workspaces?sort=-created_at&page=1&page_size=10
+```
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.entity_store.list_workspaces()
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Items per page
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[GenericSortField]` — Sort field
+
+
+
+
+
+-
+
+**filter:** `typing.Optional[str]`
+
+Query filter expression. Supports text and JSON syntaxes:
+- Text: name:"value" AND status>500 with operators : ~ > >= < <= IN NOT IN AND OR and negation prefix -
+- Object (JSON): {"name":{"$like":"value"}} with operators $eq, $like, $lt, $lte, $gt, $gte, $in, $nin, $and, $or, $not
+- Bracket notation: ?filter[name][$like]=value
+- Relationship traversal: ?filter[relationship][$exists]=true or ?filter[relationship][field]=value
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.entity_store.create_workspace(...) -> Workspace
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Create a new workspace.
+
+The creator is automatically granted Admin role on the workspace.
+By default, this endpoint waits for the Admin role to propagate before returning.
+Use `wait_role_propagation=false` to skip waiting (useful for bulk operations).
+
+Example:
+```
+POST /apis/entities/v2/workspaces
+{
+ "name": "ml-team",
+ "description": "Machine Learning Team workspace"
+}
+```
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.entity_store.create_workspace(
+ name="ml-team",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**name:** `str` — Workspace name (unique identifier). Name must start with a lowercase letter, be 2-63 characters, and contain only lowercase letters, digits, and hyphens (no consecutive hyphens, cannot end with a hyphen).
+
+
+
+
+
+-
+
+**wait_role_propagation:** `typing.Optional[bool]` — If true, wait for Admin role to propagate before returning (default: true). Set to false for bulk operations.
+
+
+
+
+
+-
+
+**description:** `typing.Optional[str]` — Optional description of the workspace
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.entity_store.get_workspace(...) -> Workspace
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get a specific workspace by ID.
+
+Example:
+```
+GET /apis/entities/v2/workspaces/ml-team
+```
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.entity_store.get_workspace(
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.entity_store.update_workspace(...) -> Workspace
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Update a workspace's description.
+
+Example:
+```
+PUT /apis/entities/v2/workspaces/ml-team
+{
+ "description": "Updated description for ML Team"
+}
+```
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.entity_store.update_workspace(
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**description:** `typing.Optional[str]` — Updated description
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.entity_store.delete_workspace(...) -> DeleteResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Delete a workspace.
+
+This marks the workspace for deletion and returns immediately. The workspace
+will no longer be accessible via the API. An asynchronous cleanup controller
+will handle deletion of all entities and external resources.
+
+Role bindings are immediately deleted to revoke access.
+
+Example:
+```
+DELETE /apis/entities/v2/workspaces/ml-team
+```
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.entity_store.delete_workspace(
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.entity_store.list_entities(...) -> EntitiesPage
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+List all entities of a specific type in the given workspace.
+
+Use workspace="-" to list entities across all workspaces the principal has
+access to.
+
+Query Parameters:
+- sort: Sort field
+- page, page_size: Pagination
+- filter: Advanced filters (JSON, text, or bracket notation)
+
+Examples:
+```
+GET /apis/entities/v2/workspaces/default/entities/customization_config?sort=-created_at
+GET /apis/entities/v2/workspaces/-/entities/customization_config # Cross-workspace query
+```
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.entity_store.list_entities(
+ workspace="workspace",
+ entity_type="entity_type",
+ sort="-created_at",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**entity_type:** `str`
+
+
+
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Items per page
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[str]` — Sort field
+
+
+
+
+
+-
+
+**filter:** `typing.Optional[str]`
+
+Query filter expression. Supports text and JSON syntaxes:
+- Text: name:"value" AND status>500 with operators : ~ > >= < <= IN NOT IN AND OR and negation prefix -
+- Object (JSON): {"name":{"$like":"value"}} with operators $eq, $like, $lt, $lte, $gt, $gte, $in, $nin, $and, $or, $not
+- Bracket notation: ?filter[name][$like]=value
+- Relationship traversal: ?filter[relationship][$exists]=true or ?filter[relationship][field]=value
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.entity_store.create_entity(...) -> Entity
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Create a new entity of the specified type in the given workspace.
+
+If name is not provided, it will be auto-generated based on the entity type.
+
+Example:
+```
+POST /apis/entities/v2/workspaces/default/entities/customization_config
+{
+ "name": "my-config",
+ "data": {
+ "target_id": "llama-2-7b",
+ "training_options": {"learning_rate": 0.01}
+ }
+}
+```
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.entity_store.create_entity(
+ workspace="workspace",
+ entity_type="entity_type",
+ data={
+ "key": "value"
+ },
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**entity_type:** `str`
+
+
+
+
+
+-
+
+**data:** `typing.Dict[str, typing.Any]` — Entity-specific data (schema is opaque to entity store, validated by client SDK)
+
+
+
+
+
+-
+
+**name:** `typing.Optional[str]` — Entity name (optional - auto-generated if not provided). Name must start with a lowercase letter, be 2-63 characters, and contain only lowercase letters, digits, and hyphens (no consecutive hyphens, cannot end with a hyphen).
+
+
+
+
+
+-
+
+**parent:** `typing.Optional[str]` — Parent entity ID for nested entities
+
+
+
+
+
+-
+
+**project:** `typing.Optional[str]` — The name of the project associated with this entity
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.entity_store.get_entity_by_name(...) -> Entity
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get a specific entity by its workspace, type, and name.
+
+Example:
+```
+GET /apis/entities/v2/workspaces/default/entities/customization_config/my-config
+```
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.entity_store.get_entity_by_name(
+ workspace="workspace",
+ entity_type="entity_type",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**entity_type:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**parent:** `typing.Optional[str]` — Parent entity ID for nested entities
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.entity_store.update_entity_by_name(...) -> Entity
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Update an entity by its name. Optionally change the entity's name.
+
+Example:
+```
+PUT /apis/entities/v2/workspaces/default/entities/customization_config/my-config
+{
+ "data": {
+ "target_id": "llama-2-7b",
+ "training_options": {"learning_rate": 0.02}
+ }
+}
+```
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.entity_store.update_entity_by_name(
+ workspace="workspace",
+ entity_type="entity_type",
+ name="name",
+ data={
+ "key": "value"
+ },
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**entity_type:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**data:** `typing.Dict[str, typing.Any]` — Updated entity-specific data
+
+
+
+
+
+-
+
+**parent:** `typing.Optional[str]` — Parent entity ID for nested entities
+
+
+
+
+
+-
+
+**new_name:** `typing.Optional[str]` — Updated entity name (optional). Name must start with a lowercase letter, be 2-63 characters, and contain only lowercase letters, digits, and hyphens (no consecutive hyphens, cannot end with a hyphen).
+
+
+
+
+
+-
+
+**project:** `typing.Optional[str]` — The name of the project associated with this entity
+
+
+
+
+
+-
+
+**expected_db_version:** `typing.Optional[int]` — Optional database version for optimistic locking. Update only succeeds if current version matches.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.entity_store.delete_entity_by_name(...) -> DeleteResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Delete an entity by its name.
+
+Example:
+```
+DELETE /apis/entities/v2/workspaces/default/entities/customization_config/my-config
+```
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.entity_store.delete_entity_by_name(
+ workspace="workspace",
+ entity_type="entity_type",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**entity_type:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**parent:** `typing.Optional[str]` — Parent entity ID for nested entities
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.entity_store.list_workspace_members(...) -> WorkspaceMemberListResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+List all members of a workspace with their roles.
+
+Returns a list of all principals with active role bindings in the workspace.
+
+Example:
+```
+GET /apis/entities/v2/workspaces/ml-team/members
+```
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.entity_store.list_workspace_members(
+ workspace="workspace",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.entity_store.add_workspace_member(...) -> WorkspaceMember
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Add a new member to the workspace with specified roles.
+
+This creates role bindings for the specified principal with the given roles.
+By default, this endpoint waits for the roles to propagate before returning.
+Use `wait_role_propagation=false` to skip waiting (useful for bulk operations).
+
+Example:
+```
+POST /apis/entities/v2/workspaces/ml-team/members
+{
+ "principal": "user@example.com",
+ "roles": ["Editor"]
+}
+```
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.entity_store.add_workspace_member(
+ workspace="workspace",
+ principal="user@example.com",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**principal:** `str` — The principal identifier (email, user ID, or group ID)
+
+
+
+
+
+-
+
+**wait_role_propagation:** `typing.Optional[bool]` — If true, wait for roles to propagate before returning (default: true). Set to false for bulk operations.
+
+
+
+
+
+-
+
+**roles:** `typing.Optional[typing.List[str]]` — List of roles to grant to the principal
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.entity_store.update_workspace_member(...) -> WorkspaceMember
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Update the roles for a workspace member.
+
+This will revoke existing roles not in the new list and add new roles.
+By default, this endpoint waits for the roles to propagate before returning.
+Use `wait_role_propagation=false` to skip waiting (useful for bulk operations).
+
+Example:
+```
+PUT /apis/entities/v2/workspaces/ml-team/members/user@example.com
+{
+ "roles": ["Viewer", "Editor"]
+}
+```
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.entity_store.update_workspace_member(
+ workspace="workspace",
+ principal_id="principal_id",
+ roles=[
+ "Viewer"
+ ],
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**principal_id:** `str`
+
+
+
+
+
+-
+
+**roles:** `typing.List[str]` — Updated list of roles for the principal
+
+
+
+
+
+-
+
+**wait_role_propagation:** `typing.Optional[bool]` — If true, wait for roles to propagate before returning (default: true). Set to false for bulk operations.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.entity_store.remove_workspace_member(...) -> DeleteResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Remove a member from the workspace by revoking all their roles.
+
+This revokes all active role bindings for the principal in the workspace.
+By default, this endpoint waits for all roles to be revoked before returning.
+Use `wait_role_propagation=false` to skip waiting (useful for bulk operations).
+
+Example:
+```
+DELETE /apis/entities/v2/workspaces/ml-team/members/user@example.com
+```
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.entity_store.remove_workspace_member(
+ workspace="workspace",
+ principal_id="principal_id",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**principal_id:** `str`
+
+
+
+
+
+-
+
+**wait_role_propagation:** `typing.Optional[bool]` — If true, wait for roles to propagate before returning (default: true). Set to false for bulk operations.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.entity_store.list_projects(...) -> ProjectsPage
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+List all projects in a workspace with pagination.
+
+Query Parameters:
+- page, page_size: Pagination
+- sort: Sort field
+- filter: Advanced filters
+
+Example:
+```
+GET /apis/entities/v2/workspaces/default/projects?sort=-created_at&page=1&page_size=10
+```
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.entity_store.list_projects(
+ workspace="workspace",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Items per page
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[ProjectSortField]` — Sort field
+
+
+
+
+
+-
+
+**filter:** `typing.Optional[str]`
+
+Query filter expression. Supports text and JSON syntaxes:
+- Text: name:"value" AND status>500 with operators : ~ > >= < <= IN NOT IN AND OR and negation prefix -
+- Object (JSON): {"name":{"$like":"value"}} with operators $eq, $like, $lt, $lte, $gt, $gte, $in, $nin, $and, $or, $not
+- Bracket notation: ?filter[name][$like]=value
+- Relationship traversal: ?filter[relationship][$exists]=true or ?filter[relationship][field]=value
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.entity_store.create_project(...) -> Project
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Create a new project in the given workspace.
+
+Example:
+```
+POST /apis/entities/v2/workspaces/default/projects
+{
+ "name": "ml-project",
+ "description": "Machine Learning project"
+}
+```
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.entity_store.create_project(
+ workspace="workspace",
+ name="ml-project",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str` — Project name (unique within workspace). Name must start with a lowercase letter, be 2-63 characters, and contain only lowercase letters, digits, and hyphens (no consecutive hyphens, cannot end with a hyphen).
+
+
+
+
+
+-
+
+**description:** `typing.Optional[str]` — Optional description of the project
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.entity_store.get_project(...) -> Project
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get a specific project by its workspace and name.
+
+Example:
+```
+GET /apis/entities/v2/workspaces/default/projects/ml-project
+```
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.entity_store.get_project(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.entity_store.update_project(...) -> Project
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Update a project's description.
+
+Example:
+```
+PUT /apis/entities/v2/workspaces/default/projects/ml-project
+{
+ "description": "Updated description for ML project"
+}
+```
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.entity_store.update_project(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**description:** `typing.Optional[str]` — Updated description
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.entity_store.delete_project(...) -> DeleteResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Delete a project.
+
+Example:
+```
+DELETE /apis/entities/v2/workspaces/default/projects/ml-project
+```
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.entity_store.delete_project(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.files.list_filesets(...) -> FilesetOutputsPage
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+List Filesets endpoint with filtering and pagination.
+
+Supports filtering by name, description, purpose, storage_type, created_at, and updated_at via query parameters.
+Returns paginated results with sorting options.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.files.list_filesets(
+ workspace="workspace",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number.
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Page size.
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[GenericSortField]` — The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+
+
+
+
+-
+
+**filter:** `typing.Optional[FilesetFilter]` — Filter filesets by name, description, purpose, storage_type, created_at, and updated_at.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.files.create_fileset(...) -> FilesetOutput
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Create a new fileset.
+
+If no storage configuration is provided, the default storage backend will be used.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.files.create_fileset(
+ workspace="workspace",
+ name="training-data-v1",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str` — The name of the fileset. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+
+
+
+
+-
+
+**description:** `typing.Optional[str]` — The description of the fileset.
+
+
+
+
+
+-
+
+**project:** `typing.Optional[str]` — The name of the project associated with this fileset.
+
+
+
+
+
+-
+
+**storage:** `typing.Optional[CreateFilesetRequestStorage]` — The storage configuration for the fileset. If not provided, uses default storage.
+
+
+
+
+
+-
+
+**purpose:** `typing.Optional[FilesetPurpose]` — The purpose of the fileset.
+
+
+
+
+
+-
+
+**metadata:** `typing.Optional[FilesetMetadataInput]` — Purpose-specific metadata. Use the purpose as the key (e.g., {dataset: {...}}).
+
+
+
+
+
+-
+
+**custom_fields:** `typing.Optional[typing.Dict[str, typing.Any]]` — Custom fields for the fileset.
+
+
+
+
+
+-
+
+**cache:** `typing.Optional[bool]` — Cache all files after creation. Only applies to external storage.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.files.retrieve_fileset(...) -> FilesetOutput
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get Fileset by Workspace and Name.
+
+Returns the details of a specific fileset identified by its workspace and name.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.files.retrieve_fileset(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.files.delete_fileset(...) -> FilesetOutput
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Delete Fileset.
+
+Permanently deletes a fileset from the platform.
+Returns metadata about the deleted fileset.
+For local storage backends, this also deletes the underlying files.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.files.delete_fileset(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.files.update_fileset_metadata(...) -> FilesetOutput
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Update Fileset Metadata.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.files.update_fileset_metadata(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**description:** `typing.Optional[str]` — The description of the fileset.
+
+
+
+
+
+-
+
+**project:** `typing.Optional[str]` — The name of the project associated with this fileset.
+
+
+
+
+
+-
+
+**purpose:** `typing.Optional[FilesetPurpose]` — The purpose of the fileset.
+
+
+
+
+
+-
+
+**metadata:** `typing.Optional[FilesetMetadataInput]` — Purpose-specific metadata. Use the purpose as the key (e.g., {dataset: {...}}).
+
+
+
+
+
+-
+
+**custom_fields:** `typing.Optional[typing.Dict[str, typing.Any]]` — Custom fields for the fileset.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.files.download_file_apis_files_v2workspaces_workspace_filesets_name_path_get(...) -> typing.Iterator[bytes]
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Download file content from a fileset.
+
+Supports HTTP Range requests for partial content retrieval (status 206).
+Returns the full file content (status 200) if no Range header is provided.
+For external resources (HuggingFace, NGC), content is cached locally on first access.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.files.download_file_apis_files_v2workspaces_workspace_filesets_name_path_get(
+ workspace="workspace",
+ name="name",
+ path="path",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**path:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.files.upload_file_apis_files_v2workspaces_workspace_filesets_name_path_put(...) -> FilesetFileOutput
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Upload file content to a fileset.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+client.files.upload_file_apis_files_v2workspaces_workspace_filesets_name_path_put(...)
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**path:** `str`
+
+
+
+
+
+-
+
+**request:** `typing.Union[bytes, typing.Iterator[bytes], typing.AsyncIterator[bytes]]` — Upload the file either as a raw octet stream.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.files.delete_file_apis_files_v2workspaces_workspace_filesets_name_path_delete(...) -> FilesetFileOutput
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Delete a specific file from a fileset.
+
+Permanently deletes the file from the storage backend.
+Returns metadata about the deleted file.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.files.delete_file_apis_files_v2workspaces_workspace_filesets_name_path_delete(
+ workspace="workspace",
+ name="name",
+ path="path",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**path:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.files.head_file_apis_files_v2workspaces_workspace_filesets_name_path_head(...) -> typing.Any
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get file metadata without downloading content.
+
+HEAD requests are often used before Range GETs to ensure the server
+supports partial downloads (e.g., DuckDB's httpfs).
+Returns Accept-Ranges, Content-Length, and Content-Type headers.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.files.head_file_apis_files_v2workspaces_workspace_filesets_name_path_head(
+ workspace="workspace",
+ name="name",
+ path="path",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**path:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.files.list_fileset_files(...) -> ListFilesetFilesResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+List Files in Fileset.
+
+Returns a list of files stored in the specified fileset.
+Optionally filter by path prefix to list files under a specific directory.
+
+Each file includes a cache_status field:
+- "not_cacheable": File is on default storage, caching not applicable
+- "cached": File exists in cache storage
+- "caching": File is currently being downloaded and cached
+- "not_cached": File not in cache, will be cached on next download
+- null: External storage, but cache status not checked (use include_cache_status=true)
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.files.list_fileset_files(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**path:** `typing.Optional[str]` — Filter files by path prefix
+
+
+
+
+
+-
+
+**include_cache_status:** `typing.Optional[bool]` — Check and return cache status for each file. When false, storage files return null for cache_status.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.otlp.upload_otlp_logs(...) -> OtelExportLogsServiceResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Upload OTLP logs to a specified fileset in JSON or Protobuf format.
+
+Supports both application/json and application/x-protobuf content types.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.otlp.upload_otlp_logs(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.otlp.query_otlp_logs(...) -> PlatformJobLogPage
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Query logs from parquet files in a fileset.
+
+This is an internal endpoint that runs DuckDB queries with direct storage
+access.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.otlp.query_otlp_logs(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**filters:** `typing.Optional[typing.Dict[str, str]]` — Key-value filters to apply to the query
+
+
+
+
+
+-
+
+**limit:** `typing.Optional[int]` — Maximum number of results to return
+
+
+
+
+
+-
+
+**page_cursor:** `typing.Optional[str]` — Cursor for pagination
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.guardrails.check(...) -> GuardrailCheckResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Chat completion for the provided conversation.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+from nvidia.guardrails import GuardrailCheckRequestMessagesItem_System
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.guardrails.check(
+ workspace="workspace",
+ model="model",
+ messages=[
+ GuardrailCheckRequestMessagesItem_System(
+ content="content",
+ )
+ ],
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**model:** `str` — The model to use for completion. Must be one of the available models.
+
+
+
+
+
+-
+
+**messages:** `typing.List[GuardrailCheckRequestMessagesItem]` — A list of messages comprising the conversation so far
+
+
+
+
+
+-
+
+**response_format:** `typing.Optional[typing.Dict[str, typing.Any]]` — Format of the response. Use {'type': 'json_object'} for JSON mode or {'type': 'json_schema', 'json_schema': {...}} for structured outputs.
+
+
+
+
+
+-
+
+**max_tokens:** `typing.Optional[int]` — The maximum number of tokens that can be generated in the chat completion.
+
+
+
+
+
+-
+
+**n:** `typing.Optional[int]` — How many chat completion choices to generate for each input message.
+
+
+
+
+
+-
+
+**stream:** `typing.Optional[bool]` — If set, partial message deltas will be sent, like in ChatGPT.
+
+
+
+
+
+-
+
+**temperature:** `typing.Optional[float]` — What sampling temperature to use, between 0 and 2.
+
+
+
+
+
+-
+
+**top_p:** `typing.Optional[float]` — An alternative to sampling with temperature, called nucleus sampling.
+
+
+
+
+
+-
+
+**stop:** `typing.Optional[GuardrailCheckRequestStop]` — Up to 4 sequences where the API will stop generating further tokens.
+
+
+
+
+
+-
+
+**frequency_penalty:** `typing.Optional[float]` — Positive values penalize new tokens based on their existing frequency in the text.
+
+
+
+
+
+-
+
+**presence_penalty:** `typing.Optional[float]` — Positive values penalize new tokens based on whether they appear in the text so far.
+
+
+
+
+
+-
+
+**function_call:** `typing.Optional[GuardrailCheckRequestFunctionCall]` — Deprecated in favor of tool_choice. 'none' means the model will not call a function and instead generates a message. 'auto' means the model can pick between generating a message or calling a function. Specifying a particular function via {'name': 'my_function'} forces the model to call that function.
+
+
+
+
+
+-
+
+**seed:** `typing.Optional[int]` — If specified, attempts to sample deterministically.
+
+
+
+
+
+-
+
+**logit_bias:** `typing.Optional[typing.Dict[str, float]]` — Modify the likelihood of specified tokens appearing in the completion. Maps token IDs (as strings) to bias values from -100 to 100.
+
+
+
+
+
+-
+
+**top_logprobs:** `typing.Optional[int]` — The number of most likely tokens to return at each token position.
+
+
+
+
+
+-
+
+**logprobs:** `typing.Optional[bool]` — Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message
+
+
+
+
+
+-
+
+**tool_choice:** `typing.Optional[GuardrailCheckRequestToolChoice]` — Controls which (if any) tool is called by the model. 'none' means no tool is called, 'auto' lets the model decide, 'required' forces a tool call.
+
+
+
+
+
+-
+
+**user:** `typing.Optional[str]` — A unique identifier representing your end-user, used by some providers for abuse monitoring.
+
+
+
+
+
+-
+
+**tools:** `typing.Optional[typing.List[typing.Dict[str, typing.Any]]]` — A list of tools the model may call. Each tool is an object with a 'type' field and a 'function' definition.
+
+
+
+
+
+-
+
+**ignore_eos:** `typing.Optional[bool]` — Ignore the eos when running
+
+
+
+
+
+-
+
+**reasoning_effort:** `typing.Optional[str]` — Constrains effort on reasoning for reasoning models. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.
+
+
+
+
+
+-
+
+**max_completion_tokens:** `typing.Optional[int]` — An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens. Preferred over max_tokens for reasoning models.
+
+
+
+
+
+-
+
+**stream_options:** `typing.Optional[typing.Dict[str, bool]]` — Options for streaming response. Only set this when stream=True. Supports include_usage to receive token usage in the final stream chunk.
+
+
+
+
+
+-
+
+**vision:** `typing.Optional[bool]` — Whether this is a vision-capable request with image inputs.
+
+
+
+
+
+-
+
+**guardrails:** `typing.Optional[GuardrailsDataInput]` — Guardrails specific options for the request.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.guardrails.list_guardrail_configs(...) -> GuardrailConfigsPage
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+List available guardrail configs.
+
+Lists guardrail configs for a specific workspace.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.guardrails.list_guardrail_configs(
+ workspace="workspace",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number.
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Page size.
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[GenericSortField]` — The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+
+
+
+
+-
+
+**filter:** `typing.Optional[GuardrailConfigFilter]` — Filter guardrail configs by name, description, project, created_at, and updated_at.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.guardrails.create_config(...) -> GuardrailConfig
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Create a new guardrail config.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.guardrails.create_config(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str` — The name of the guardrail config
+
+
+
+
+
+-
+
+**description:** `typing.Optional[str]` — Description of the guardrail config
+
+
+
+
+
+-
+
+**data:** `typing.Optional[typing.Dict[str, typing.Any]]` — Guardrail configuration data
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.guardrails.get_guardrail_config(...) -> GuardrailConfig
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get info about a guardrail configuration.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.guardrails.get_guardrail_config(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.guardrails.delete_config(...) -> DeleteResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Delete a guardrail config.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.guardrails.delete_config(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.guardrails.update_config(...) -> GuardrailConfig
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Update model metadata. If the request body has an empty field,
+keep the old value.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.guardrails.update_config(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**description:** `typing.Optional[str]` — Description of the guardrail config
+
+
+
+
+
+-
+
+**data:** `typing.Optional[typing.Dict[str, typing.Any]]` — Guardrail configuration data
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.inference_gateway.gateway_proxy_get(...) -> typing.Any
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Proxy requests to model entity inference endpoints.
+
+All inference requests must resolve to a `VirtualModel`. The platform's
+provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+for every served model entity (named after the entity, with
+`default_model_entity` set to the entity ref) so this is the typical case;
+operators can also create custom VirtualModels for routing, plugin chains,
+LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+return `404`.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.inference_gateway.gateway_proxy_get(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**trailing_uri:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.inference_gateway.gateway_proxy_post(...) -> typing.Dict[str, typing.Any]
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Proxy requests to model entity inference endpoints.
+
+All inference requests must resolve to a `VirtualModel`. The platform's
+provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+for every served model entity (named after the entity, with
+`default_model_entity` set to the entity ref) so this is the typical case;
+operators can also create custom VirtualModels for routing, plugin chains,
+LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+return `404`.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.inference_gateway.gateway_proxy_post(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ request={
+ "key": "value"
+ },
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**trailing_uri:** `str`
+
+
+
+
+
+-
+
+**request:** `typing.Dict[str, typing.Any]`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.inference_gateway.gateway_proxy_put(...) -> typing.Dict[str, typing.Any]
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Proxy requests to model entity inference endpoints.
+
+All inference requests must resolve to a `VirtualModel`. The platform's
+provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+for every served model entity (named after the entity, with
+`default_model_entity` set to the entity ref) so this is the typical case;
+operators can also create custom VirtualModels for routing, plugin chains,
+LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+return `404`.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.inference_gateway.gateway_proxy_put(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ request={
+ "key": "value"
+ },
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**trailing_uri:** `str`
+
+
+
+
+
+-
+
+**request:** `typing.Dict[str, typing.Any]`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.inference_gateway.gateway_proxy_delete(...) -> typing.Any
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Proxy requests to model entity inference endpoints.
+
+All inference requests must resolve to a `VirtualModel`. The platform's
+provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+for every served model entity (named after the entity, with
+`default_model_entity` set to the entity ref) so this is the typical case;
+operators can also create custom VirtualModels for routing, plugin chains,
+LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+return `404`.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.inference_gateway.gateway_proxy_delete(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**trailing_uri:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.inference_gateway.gateway_proxy_patch(...) -> typing.Dict[str, typing.Any]
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Proxy requests to model entity inference endpoints.
+
+All inference requests must resolve to a `VirtualModel`. The platform's
+provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+for every served model entity (named after the entity, with
+`default_model_entity` set to the entity ref) so this is the typical case;
+operators can also create custom VirtualModels for routing, plugin chains,
+LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+return `404`.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.inference_gateway.gateway_proxy_patch(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ request={
+ "key": "value"
+ },
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**trailing_uri:** `str`
+
+
+
+
+
+-
+
+**request:** `typing.Dict[str, typing.Any]`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.inference_gateway.openai_proxy_list_models(...) -> OpenAiListModelsResp
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+This endpoint aggregates models from all model entities and returns them
+in OpenAI's list models format. Each model ID is the model entity identifier
+in format workspace/model_entity_name.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.inference_gateway.openai_proxy_list_models(
+ workspace="workspace",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.inference_gateway.openai_proxy_get_model(...) -> OpenAiModelResp
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Retrieve information about a specific OpenAI-compatible model.
+Workspace is always taken from the URL path; name may be model_entity_name
+or workspace/model_entity_name (workspace prefix is ignored).
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.inference_gateway.openai_proxy_get_model(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.inference_gateway.openai_proxy_get(...) -> typing.Any
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Proxy requests to OpenAI-compatible inference endpoints.
+
+All inference requests must resolve to a `VirtualModel`. The platform's
+provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+for every served model entity (named after the entity, with
+`default_model_entity` set to the entity ref) so this is the typical case;
+operators can also create custom VirtualModels for routing, plugin chains,
+LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+return `404`.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.inference_gateway.openai_proxy_get(
+ workspace="workspace",
+ trailing_uri="trailing_uri",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**trailing_uri:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.inference_gateway.openai_proxy_post(...) -> typing.Dict[str, typing.Any]
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Proxy requests to OpenAI-compatible inference endpoints.
+
+All inference requests must resolve to a `VirtualModel`. The platform's
+provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+for every served model entity (named after the entity, with
+`default_model_entity` set to the entity ref) so this is the typical case;
+operators can also create custom VirtualModels for routing, plugin chains,
+LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+return `404`.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.inference_gateway.openai_proxy_post(
+ workspace="workspace",
+ trailing_uri="trailing_uri",
+ request={
+ "key": "value"
+ },
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**trailing_uri:** `str`
+
+
+
+
+
+-
+
+**request:** `typing.Dict[str, typing.Any]`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.inference_gateway.openai_proxy_put(...) -> typing.Dict[str, typing.Any]
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Proxy requests to OpenAI-compatible inference endpoints.
+
+All inference requests must resolve to a `VirtualModel`. The platform's
+provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+for every served model entity (named after the entity, with
+`default_model_entity` set to the entity ref) so this is the typical case;
+operators can also create custom VirtualModels for routing, plugin chains,
+LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+return `404`.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.inference_gateway.openai_proxy_put(
+ workspace="workspace",
+ trailing_uri="trailing_uri",
+ request={
+ "key": "value"
+ },
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**trailing_uri:** `str`
+
+
+
+
+
+-
+
+**request:** `typing.Dict[str, typing.Any]`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.inference_gateway.openai_proxy_delete(...) -> typing.Any
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Proxy requests to OpenAI-compatible inference endpoints.
+
+All inference requests must resolve to a `VirtualModel`. The platform's
+provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+for every served model entity (named after the entity, with
+`default_model_entity` set to the entity ref) so this is the typical case;
+operators can also create custom VirtualModels for routing, plugin chains,
+LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+return `404`.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.inference_gateway.openai_proxy_delete(
+ workspace="workspace",
+ trailing_uri="trailing_uri",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**trailing_uri:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.inference_gateway.openai_proxy_patch(...) -> typing.Dict[str, typing.Any]
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Proxy requests to OpenAI-compatible inference endpoints.
+
+All inference requests must resolve to a `VirtualModel`. The platform's
+provider reconciler auto-creates an implicit `autoprovisioned` VirtualModel
+for every served model entity (named after the entity, with
+`default_model_entity` set to the entity ref) so this is the typical case;
+operators can also create custom VirtualModels for routing, plugin chains,
+LoRA escape-hatches, etc. Requests for which no VirtualModel can be found
+return `404`.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.inference_gateway.openai_proxy_patch(
+ workspace="workspace",
+ trailing_uri="trailing_uri",
+ request={
+ "key": "value"
+ },
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**trailing_uri:** `str`
+
+
+
+
+
+-
+
+**request:** `typing.Dict[str, typing.Any]`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.inference_gateway.provider_proxy_get(...) -> typing.Any
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Proxy requests to provider inference endpoints.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.inference_gateway.provider_proxy_get(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**trailing_uri:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.inference_gateway.provider_proxy_post(...) -> typing.Dict[str, typing.Any]
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Proxy requests to provider inference endpoints.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.inference_gateway.provider_proxy_post(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ request={
+ "key": "value"
+ },
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**trailing_uri:** `str`
+
+
+
+
+
+-
+
+**request:** `typing.Dict[str, typing.Any]`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.inference_gateway.provider_proxy_put(...) -> typing.Dict[str, typing.Any]
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Proxy requests to provider inference endpoints.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.inference_gateway.provider_proxy_put(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ request={
+ "key": "value"
+ },
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**trailing_uri:** `str`
+
+
+
+
+
+-
+
+**request:** `typing.Dict[str, typing.Any]`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.inference_gateway.provider_proxy_delete(...) -> typing.Any
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Proxy requests to provider inference endpoints.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.inference_gateway.provider_proxy_delete(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**trailing_uri:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.inference_gateway.provider_proxy_patch(...) -> typing.Dict[str, typing.Any]
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Proxy requests to provider inference endpoints.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.inference_gateway.provider_proxy_patch(
+ workspace="workspace",
+ name="name",
+ trailing_uri="trailing_uri",
+ request={
+ "key": "value"
+ },
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**trailing_uri:** `str`
+
+
+
+
+
+-
+
+**request:** `typing.Dict[str, typing.Any]`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.inference_gateway.provider_ready(...) -> typing.Dict[str, typing.Any]
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Check if a model provider is registered in the gateway's cache.
+
+This is a lightweight endpoint that only checks the gateway's internal state,
+without making any requests to the actual provider backend. Use this to verify
+the gateway is ready to route requests to a provider after deployment.
+
+Returns:
+ 200 OK with provider info if the provider is registered
+ 404 Not Found if the provider is not yet in the gateway's cache
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.inference_gateway.provider_ready(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.virtual_models.list_virtual_models(...) -> VirtualModelsPage
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+List VirtualModels for the given workspace.
+
+Use ``workspace=-`` to list across all workspaces accessible to the caller.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.virtual_models.list_virtual_models(
+ workspace="workspace",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number (1-indexed).
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Number of results per page.
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[str]` — Sort field. Prefix with ``-`` for descending order.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.virtual_models.create_virtual_model(...) -> VirtualModel
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Create a new VirtualModel in the given workspace.
+
+A VirtualModel defines an ordered middleware pipeline that IGW executes
+when an inference request arrives with ``model: "workspace/name"`` matching
+this entity.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.virtual_models.create_virtual_model(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str` — Name of the virtual model within the workspace. Must be unique per workspace.
+
+
+
+
+
+-
+
+**default_model_entity:** `typing.Optional[str]` — Model entity to route to, in "workspace/name" format. Written into request["model"] before the request middleware pipeline runs. If omitted, a request middleware plugin must handle backend routing itself. Set to null to clear an existing value.
+
+
+
+
+
+-
+
+**autoprovisioned:** `typing.Optional[bool]` — Marks this VirtualModel as controller-managed. The Models controller will delete it once no ModelProvider serves the matching entity. Setting this manually opts the VirtualModel into that cleanup behavior.
+
+
+
+
+
+-
+
+**models:** `typing.Optional[typing.List[VirtualModelInferenceConfig]]` — Model entity references used by this VirtualModel. A per-entry backend_format overrides the referenced ModelEntity backend_format when IGW resolves the backend format for a request.
+
+
+
+
+
+-
+
+**request_middleware:** `typing.Optional[typing.List[MiddlewareCall]]` — Ordered list of middleware plugins applied before proxying to the backend. Each entry is a MiddlewareCall with a "name" (plugin identifier) and optional "config_type" and "config_id" fields that reference a stored plugin configuration.
+
+
+
+
+
+-
+
+**response_middleware:** `typing.Optional[typing.List[MiddlewareCall]]` — Ordered list of middleware plugins applied after the backend response is received, before returning it to the caller.
+
+
+
+
+
+-
+
+**post_response_middleware:** `typing.Optional[typing.List[MiddlewareCall]]` — Ordered list of middleware plugins invoked after the response has been returned to the caller. Intended for fire-and-forget work (logging, analytics) that must not block or modify the response.
+
+
+
+
+
+-
+
+**override_proxy:** `typing.Optional[str]` — Plugin-provided proxy implementation for IGW to use instead of its default aiohttp proxy. Format: "plugin-name.proxy-name". Leave unset to use the default IGW proxy. Set to null to clear an existing value.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.virtual_models.get_virtual_model(...) -> VirtualModel
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get a VirtualModel by workspace and name.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.virtual_models.get_virtual_model(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.virtual_models.delete_virtual_model(...)
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Permanently delete a VirtualModel.
+
+This does not affect any in-flight requests already being routed through
+this VirtualModel. IGW's model cache is refreshed on its next polling cycle.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.virtual_models.delete_virtual_model(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.virtual_models.update_virtual_model(...) -> VirtualModel
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Partially update a VirtualModel.
+
+Only fields present in the request body are modified. Fields absent from
+the request body retain their current values.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.virtual_models.update_virtual_model(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**default_model_entity:** `typing.Optional[str]` — Model entity to route to, in "workspace/name" format. Written into request["model"] before the request middleware pipeline runs. If omitted, a request middleware plugin must handle backend routing itself. Set to null to clear an existing value.
+
+
+
+
+
+-
+
+**autoprovisioned:** `typing.Optional[bool]` — Marks this VirtualModel as controller-managed. The Models controller will delete it once no ModelProvider serves the matching entity. Setting this manually opts the VirtualModel into that cleanup behavior.
+
+
+
+
+
+-
+
+**models:** `typing.Optional[typing.List[VirtualModelInferenceConfig]]` — Model entity references used by this VirtualModel. A per-entry backend_format overrides the referenced ModelEntity backend_format when IGW resolves the backend format for a request.
+
+
+
+
+
+-
+
+**request_middleware:** `typing.Optional[typing.List[MiddlewareCall]]` — Ordered list of middleware plugins applied before proxying to the backend. Each entry is a MiddlewareCall with a "name" (plugin identifier) and optional "config_type" and "config_id" fields that reference a stored plugin configuration.
+
+
+
+
+
+-
+
+**response_middleware:** `typing.Optional[typing.List[MiddlewareCall]]` — Ordered list of middleware plugins applied after the backend response is received, before returning it to the caller.
+
+
+
+
+
+-
+
+**post_response_middleware:** `typing.Optional[typing.List[MiddlewareCall]]` — Ordered list of middleware plugins invoked after the response has been returned to the caller. Intended for fire-and-forget work (logging, analytics) that must not block or modify the response.
+
+
+
+
+
+-
+
+**override_proxy:** `typing.Optional[str]` — Plugin-provided proxy implementation for IGW to use instead of its default aiohttp proxy. Format: "plugin-name.proxy-name". Leave unset to use the default IGW proxy. Set to null to clear an existing value.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.annotations.list_annotations(...) -> AnnotationsPage
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.annotations.list_annotations(
+ workspace="workspace",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number.
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Page size.
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[AnnotationSortField]`
+
+
+
+
+
+-
+
+**filter:** `typing.Optional[AnnotationFilter]` — Filter annotations by span_id, session_id, kind, name, created_by, and created_at range.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.annotations.create_annotation(...) -> Annotation
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi, AnnotationInput_Feedback
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.annotations.create_annotation(
+ workspace="workspace",
+ request=AnnotationInput_Feedback(
+ session_id="session_id",
+ value="positive",
+ ),
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**request:** `AnnotationInput`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.annotations.get_annotation(...) -> Annotation
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.annotations.get_annotation(
+ workspace="workspace",
+ annotation_id="annotation_id",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**annotation_id:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.annotations.delete_annotation(...)
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.annotations.delete_annotation(
+ workspace="workspace",
+ annotation_id="annotation_id",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**annotation_id:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.evaluator_results.list_evaluator_results(...) -> EvaluatorResultsPage
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.evaluator_results.list_evaluator_results(
+ workspace="workspace",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number.
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Page size.
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[EvaluatorResultSortField]`
+
+
+
+
+
+-
+
+**filter:** `typing.Optional[EvaluatorResultFilter]` — Filter evaluator results by span_id, session_id, name, data_type, created_by, value range, and created_at range.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.evaluator_results.create_evaluator_result(...) -> EvaluatorResult
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.evaluator_results.create_evaluator_result(
+ workspace="workspace",
+ span_id="span_id",
+ session_id="session_id",
+ name="name",
+ data_type="NUMERIC",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**span_id:** `str` — Target span id. Not validated against existing spans (loose target policy).
+
+
+
+
+
+-
+
+**session_id:** `str` — Session id the target span belongs to. Denormalized so session-scoped reads stay fast.
+
+
+
+
+
+-
+
+**name:** `str` — Evaluator / metric identity (e.g. 'faithfulness/v1').
+
+
+
+
+
+-
+
+**data_type:** `EvaluatorResultDataType` — Discriminator for which of value / string_value carries the payload.
+
+
+
+
+
+-
+
+**value:** `typing.Optional[float]` — Numeric value. Required when data_type is NUMERIC or BOOLEAN (0|1).
+
+
+
+
+
+-
+
+**string_value:** `typing.Optional[str]` — String value. Required when data_type is CATEGORICAL or TEXT.
+
+
+
+
+
+-
+
+**comment:** `typing.Optional[str]` — Free-text rationale or explanation.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.evaluator_results.get_evaluator_result(...) -> EvaluatorResult
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.evaluator_results.get_evaluator_result(
+ workspace="workspace",
+ evaluator_result_id="evaluator_result_id",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**evaluator_result_id:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.evaluator_results.list_evaluator_results_for_span(...) -> typing.List[EvaluatorResult]
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.evaluator_results.list_evaluator_results_for_span(
+ workspace="workspace",
+ span_id="span_id",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**span_id:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.experiment_groups.list_experiment_groups(...) -> ExperimentGroupResponsesPage
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.experiment_groups.list_experiment_groups(
+ workspace="workspace",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number.
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Page size.
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[ListExperimentGroupsApisIntakeV2WorkspacesWorkspaceExperimentGroupsGetRequestSort]` — Sort field; prefix with '-' for descending.
+
+
+
+
+
+-
+
+**filter:** `typing.Optional[ExperimentGroupFilter]` — Filter experiment groups by name.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.experiment_groups.create_experiment_group(...) -> ExperimentGroupResponse
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.experiment_groups.create_experiment_group(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**request:** `ExperimentGroupRequest`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.experiment_groups.get_experiment_group(...) -> ExperimentGroupResponse
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.experiment_groups.get_experiment_group(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.experiment_groups.update_experiment_group(...) -> ExperimentGroupResponse
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.experiment_groups.update_experiment_group(
+ workspace="workspace",
+ name_="name",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request:** `ExperimentGroupRequest`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.experiment_groups.delete_experiment_group(...)
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.experiment_groups.delete_experiment_group(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.experiments.list_experiments(...) -> ExperimentResponsesPage
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.experiments.list_experiments(
+ workspace="workspace",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number.
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Page size.
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[ListExperimentsApisIntakeV2WorkspacesWorkspaceExperimentsGetRequestSort]` — Sort field; prefix with '-' for descending.
+
+
+
+
+
+-
+
+**filter:** `typing.Optional[ExperimentFilter]` — Filter experiments by name, experiment_group_id, agent_name, agent_version, dataset_name, dataset_version, created_by, created_at, or updated_at.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.experiments.create_experiment(...) -> ExperimentResponse
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.experiments.create_experiment(
+ workspace="workspace",
+ name="name",
+ agent_name="agent_name",
+ agent_version="agent_version",
+ dataset_name="dataset_name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**request:** `ExperimentRequest`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.experiments.get_experiment(...) -> ExperimentResponse
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.experiments.get_experiment(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.experiments.update_experiment(...) -> ExperimentResponse
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.experiments.update_experiment(
+ workspace="workspace",
+ name_="name",
+ name="name",
+ agent_name="agent_name",
+ agent_version="agent_version",
+ dataset_name="dataset_name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request:** `ExperimentRequest`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.experiments.delete_experiment(...)
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.experiments.delete_experiment(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.experiments.list_experiment_sessions(...) -> ExperimentSessionResponsesPage
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.experiments.list_experiment_sessions(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number.
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Page size.
+
+
+
+
+
+-
+
+**filter:** `typing.Optional[ExperimentSessionFilter]` — Filter sessions by test_case_id and status.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.ingest.ingest_atif(...)
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi, AtifAgent
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.ingest.ingest_atif(
+ workspace="workspace",
+ schema_version="ATIF-v1.0",
+ agent=AtifAgent(
+ name="name",
+ version="version",
+ ),
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**schema_version:** `AtifIngestRequestSchemaVersion`
+
+
+
+
+
+-
+
+**agent:** `AtifAgent`
+
+
+
+
+
+-
+
+**experiment_context:** `typing.Optional[ExperimentContext]`
+
+
+
+
+
+-
+
+**evaluation_context:** `typing.Optional[EvaluationContext]` — Deprecated. Use experiment_context; when both are sent, experiment_context takes precedence.
+
+
+
+
+
+-
+
+**session_id:** `typing.Optional[str]`
+
+
+
+
+
+-
+
+**final_metrics:** `typing.Optional[AtifFinalMetrics]`
+
+
+
+
+
+-
+
+**continued_trajectory_ref:** `typing.Optional[str]`
+
+
+
+
+
+-
+
+**notes:** `typing.Optional[str]`
+
+
+
+
+
+-
+
+**extra:** `typing.Optional[typing.Dict[str, typing.Any]]`
+
+
+
+
+
+-
+
+**steps:** `typing.Optional[typing.List[AtifStep]]`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.ingest.ingest_chat_completion(...) -> ChatCompletionsIngestResponse
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi, CapturedChatCompletionsRequest, CapturedChatMessage
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.ingest.ingest_chat_completion(
+ workspace="workspace",
+ request=CapturedChatCompletionsRequest(
+ messages=[
+ CapturedChatMessage(
+ role="user",
+ ),
+ CapturedChatMessage(
+ role="user",
+ )
+ ],
+ model="model",
+ ),
+ response={"key": "value"},
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**request:** `CapturedChatCompletionsRequest`
+
+
+
+
+
+-
+
+**response:** `CapturedChatCompletionsResponse`
+
+
+
+
+
+-
+
+**experiment_context:** `typing.Optional[ExperimentContext]`
+
+
+
+
+
+-
+
+**evaluation_context:** `typing.Optional[EvaluationContext]` — Deprecated. Use experiment_context; when both are sent, experiment_context takes precedence.
+
+
+
+
+
+-
+
+**session_id:** `typing.Optional[str]` — Groups related chat-completions calls without forcing them into the same trace.
+
+
+
+
+
+-
+
+**trace_id:** `typing.Optional[str]` — Opt into joining an existing trace built via OTel or ATIF. This is not a grouping mechanism for chat-completions calls; use session_id to group related calls.
+
+
+
+
+
+-
+
+**provider:** `typing.Optional[str]`
+
+
+
+
+
+-
+
+**cost_usd:** `typing.Optional[float]` — Total estimated cost of this model call in USD. This matches ATIF step metrics; Intake stores it as semantic cost_total_usd on spans.
+
+
+
+
+
+-
+
+**cost_input_usd:** `typing.Optional[float]` — Estimated input-token cost of this model call in USD.
+
+
+
+
+
+-
+
+**cost_output_usd:** `typing.Optional[float]` — Estimated output-token cost of this model call in USD.
+
+
+
+
+
+-
+
+**cost_details:** `typing.Optional[typing.Dict[str, float]]` — Additional estimated cost breakdown fields in USD.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.ingest.ingest_otlp_traces(...) -> IngestResponse
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.ingest.ingest_otlp_traces(
+ workspace="workspace",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.spans.list_spans(...) -> SpansPage
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.spans.list_spans(
+ workspace="workspace",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number.
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Page size.
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[SpanSortField]`
+
+
+
+
+
+-
+
+**mode:** `typing.Optional[ListSpansApisIntakeV2WorkspacesWorkspaceSpansGetRequestMode]`
+
+
+
+
+
+-
+
+**filter:** `typing.Optional[SpanFilter]` — Filter spans by session_id, trace_id, parent_span_id, project, evaluation context fields, source, kind, status, model, tool_name, provider, agent_id, agent_name, prompt_name, prompt_version, and started_at.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.spans.get_span(...) -> Span
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.spans.get_span(
+ workspace="workspace",
+ span_id="span_id",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**span_id:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.traces.list_traces(...) -> TracesPage
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.traces.list_traces(
+ workspace="workspace",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number.
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Page size.
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[TraceSortField]`
+
+
+
+
+
+-
+
+**mode:** `typing.Optional[ListTracesApisIntakeV2WorkspacesWorkspaceTracesGetRequestMode]` — Use summary for root-span trace fields only, or detailed to include token, cost, and span-count rollups.
+
+
+
+
+
+-
+
+**filter:** `typing.Optional[TraceFilter]` — Filter root-span-backed traces by id, session_id, rolled-up status, root span started_at, and root-span evaluation context fields.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.traces.get_trace(...) -> Trace
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.traces.get_trace(
+ workspace="workspace",
+ id="id",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**id:** `str`
+
+
+
+
+
+-
+
+**mode:** `typing.Optional[GetTraceApisIntakeV2WorkspacesWorkspaceTracesIdGetRequestMode]` — Use summary for root-span trace fields only, or detailed to include token, cost, and span-count rollups.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.get_execution_profiles() -> typing.List[GetExecutionProfilesApisJobsV2ExecutionProfilesGetResponseItem]
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get all currently configured execution profiles.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.get_execution_profiles()
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.list_jobs(...) -> PlatformJobResponsesPage
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+List platform jobs with filtering and pagination.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.list_jobs(
+ workspace="workspace",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number.
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Page size.
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[PlatformJobSortField]` — The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+
+
+
+
+-
+
+**filter:** `typing.Optional[PlatformJobsListFilter]` — Filter jobs by workspace, project, name, status, source, created_at, and updated_at.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.create_job(...) -> PlatformJobResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Create a new platform job.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi, PlatformJobSpecInput, PlatformJobStepSpecInput, PlatformJobStepSpecInputExecutor_Cpu, ContainerSpec
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.create_job(
+ workspace="workspace",
+ spec={
+ "key": "value"
+ },
+ platform_spec=PlatformJobSpecInput(
+ steps=[
+ PlatformJobStepSpecInput(
+ name="preprocess",
+ executor=PlatformJobStepSpecInputExecutor_Cpu(
+ container=ContainerSpec(
+ image="image",
+ ),
+ ),
+ )
+ ],
+ ),
+ source="source",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**spec:** `typing.Dict[str, typing.Any]`
+
+
+
+
+
+-
+
+**platform_spec:** `PlatformJobSpecInput`
+
+
+
+
+
+-
+
+**source:** `str`
+
+
+
+
+
+-
+
+**name:** `typing.Optional[str]`
+
+
+
+
+
+-
+
+**description:** `typing.Optional[str]`
+
+
+
+
+
+-
+
+**project:** `typing.Optional[str]`
+
+
+
+
+
+-
+
+**ownership:** `typing.Optional[typing.Dict[str, typing.Any]]`
+
+
+
+
+
+-
+
+**custom_fields:** `typing.Optional[typing.Dict[str, typing.Any]]`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.get_job_result(...) -> PlatformJobResultResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get a specific job result.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.get_job_result(
+ workspace="workspace",
+ job="job",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**job:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.create_job_result(...) -> PlatformJobResultResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Create a new result for a job.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.create_job_result(
+ workspace="workspace",
+ job="job",
+ name="name",
+ artifact_url="artifact_url",
+ artifact_storage_type="fileset",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**job:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**artifact_url:** `str`
+
+
+
+
+
+-
+
+**artifact_storage_type:** `FileStorageType`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.download_job_result(...) -> typing.Iterator[bytes]
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Download a job result file.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.download_job_result(
+ workspace="workspace",
+ job="job",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**job:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.get_job_step(...) -> PlatformJobStep
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get a specific job step.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.get_job_step(
+ workspace="workspace",
+ job="job",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**job:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.update_job_step_status(...) -> PlatformJobStep
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Update a job step status.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.update_job_step_status(
+ workspace="workspace",
+ job="job",
+ name="name",
+ status="created",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**job:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**status:** `PlatformJobStatus` — The new status to set for the job.
+
+
+
+
+
+-
+
+**status_details:** `typing.Optional[typing.Dict[str, typing.Any]]` — Optional status details related to the status update.
+
+
+
+
+
+-
+
+**error_details:** `typing.Optional[typing.Dict[str, typing.Any]]` — Optional error details related to the status update.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.list_job_step_tasks(...) -> PlatformJobListTaskResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+List tasks for a job step.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.list_job_step_tasks(
+ workspace="workspace",
+ job="job",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**job:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.get_job_step_task(...) -> PlatformJobTask
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get a specific job step task.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.get_job_step_task(
+ workspace="workspace",
+ job="job",
+ step="step",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**job:** `str`
+
+
+
+
+
+-
+
+**step:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.update_job_step_task(...) -> PlatformJobTask
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Update a job step task.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.update_job_step_task(
+ workspace="workspace",
+ job="job",
+ step="step",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**job:** `str`
+
+
+
+
+
+-
+
+**step:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**status:** `typing.Optional[PlatformJobStatus]`
+
+
+
+
+
+-
+
+**status_details:** `typing.Optional[typing.Dict[str, typing.Any]]`
+
+
+
+
+
+-
+
+**error_details:** `typing.Optional[typing.Dict[str, typing.Any]]`
+
+
+
+
+
+-
+
+**error_stack:** `typing.Optional[str]`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.get_job(...) -> PlatformJobResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get a platform job by name.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.get_job(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.delete_job(...)
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Delete a platform job.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.delete_job(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.cancel_job(...) -> PlatformJobResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Cancel a platform job.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.cancel_job(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.page_job_logs(...) -> PlatformJobLogPage
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get paginated logs for a platform job.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.page_job_logs(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**limit:** `typing.Optional[int]` — Maximum number of logs to return
+
+
+
+
+
+-
+
+**page_cursor:** `typing.Optional[str]` — Page cursor
+
+
+
+
+
+-
+
+**attempt_id:** `typing.Optional[int]` — Filter logs by job attempt ID
+
+
+
+
+
+-
+
+**step_id:** `typing.Optional[str]` — Filter logs by step name
+
+
+
+
+
+-
+
+**task_id:** `typing.Optional[str]` — Filter logs by task ID
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.pause_job(...) -> PlatformJobResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Pause a platform job.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.pause_job(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.list_job_results(...) -> PlatformJobListResultResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+List results for a job.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.list_job_results(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[PlatformJobSortField]` — The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.resume_job(...) -> PlatformJobResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Resume a paused platform job.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.resume_job(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.get_job_status(...) -> PlatformJobStatusResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get the status of a platform job.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.get_job_status(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.update_job_status_details(...) -> typing.Any
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Update the status details of a platform job.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.update_job_status_details(
+ workspace="workspace",
+ name="name",
+ request={
+ "key": "value"
+ },
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request:** `typing.Dict[str, typing.Any]`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.jobs.list_steps(...) -> PlatformJobStepWithContextsPage
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+List job steps with pagination and filtering.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.jobs.list_steps(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number.
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Page size.
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[PlatformJobSortField]` — The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+
+
+
+
+-
+
+**filter:** `typing.Optional[PlatformJobStepsListFilter]` — Filter steps by job, status, and source.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.adapters.list_adapters(...) -> AdaptersPage
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.adapters.list_adapters(
+ workspace="workspace",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number.
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Page size.
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[str]` — The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+
+
+
+
+-
+
+**filter:** `typing.Optional[AdapterEntityFilter]` — Filter adapters by name, model (parent model ref string, stored on the adapter), description, fileset, finetuning_type, enabled, created_at, and updated_at.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.adapters.create_adapter(...) -> Adapter
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Create an adapter under a base model specified by the "model" field in the body.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.adapters.create_adapter(
+ workspace="workspace",
+ name="lora-adapter-v1",
+ fileset="fileset",
+ finetuning_type="lora_merged",
+ model="llama-3-8b-instruct",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str` — Name of the adapter. Name must be unique in the workspace. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+
+
+
+
+-
+
+**fileset:** `str` — Location where adapter files are stored - expected format {workspace}/{fileset_name}
+
+
+
+
+
+-
+
+**finetuning_type:** `FinetuningType` — Type of finetuning (LORA, P_TUNING, etc.)
+
+
+
+
+
+-
+
+**model:** `str`
+
+Base model entity.
+ Use `{workspace}/{model_name}` to reference a model in any workspace, or a single `{model_name}` resolved in the path workspace. A single name (2-63 characters) or 'workspace/model_name' where each segment is a valid name (lowercase, digits, hyphens, and temporarily @ . + _; no leading/trailing or consecutive hyphens). If one slash, both sides must be non-empty.
+
+
+
+
+
+-
+
+**description:** `typing.Optional[str]` — Optional description of the adapter
+
+
+
+
+
+-
+
+**enabled:** `typing.Optional[bool]` — Whether to make this adapter available for inference post training
+
+
+
+
+
+-
+
+**lora_config:** `typing.Optional[Lora]` — Lora configuration specifics
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.adapters.get_adapter(...) -> Adapter
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.adapters.get_adapter(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.adapters.delete_adapter(...)
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.adapters.delete_adapter(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.adapters.update_adapter(...) -> Adapter
+
+-
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.adapters.update_adapter(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request:** `UpdateAdapterRequest`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_deployment_configs.list_deployment_configs(...) -> ModelDeploymentConfigsPage
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+List ModelDeploymentConfigs for a specific workspace.
+Returns only the latest version of each config.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_deployment_configs.list_deployment_configs(
+ workspace="workspace",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number.
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Page size.
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[str]` — The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+
+
+
+
+-
+
+**filter:** `typing.Optional[ModelDeploymentConfigFilter]` — Filter deployment configs by workspace, project, model_entity_id, name, description, created_at, and updated_at.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_deployment_configs.create_deployment_config(...) -> ModelDeploymentConfig
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Create a new ModelDeploymentConfig (version 1).
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi, ModelDeploymentConfigModelSpec, ContainerExecutorConfig
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_deployment_configs.create_deployment_config(
+ workspace="workspace",
+ name="nim-config-v1",
+ engine="nim",
+ model_spec=ModelDeploymentConfigModelSpec(),
+ executor_config=ContainerExecutorConfig(
+ gpu=1,
+ ),
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str` — Name of the deployment configuration. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+
+
+
+
+-
+
+**engine:** `Engine` — Inference engine selecting the compiler path (nim/vllm/generic)
+
+
+
+
+
+-
+
+**model_spec:** `ModelDeploymentConfigModelSpec` — What model to serve and how -- independent of the executor it runs on
+
+
+
+
+
+-
+
+**executor_config:** `ContainerExecutorConfig` — Compute + container settings for the executor the deployment runs on
+
+
+
+
+
+-
+
+**project:** `typing.Optional[str]` — The URN of the project associated with this deployment configuration
+
+
+
+
+
+-
+
+**description:** `typing.Optional[str]` — Optional description of the deployment configuration
+
+
+
+
+
+-
+
+**model_entity_id:** `typing.Optional[str]` — Optional reference to the base model entity ID for this deployment
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_deployment_configs.get_deployment_config_version(...) -> ModelDeploymentConfig
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get a specific version of a ModelDeploymentConfig.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_deployment_configs.get_deployment_config_version(
+ workspace="workspace",
+ config="config",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**config:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_deployment_configs.delete_deployment_config_version(...)
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Delete a specific version of a ModelDeploymentConfig.
+
+This operation will fail with 409 Conflict if any ModelDeployments currently
+reference this specific version and are not in DELETED status. Delete or wait for
+dependent deployments to reach DELETED status before deleting the config version.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_deployment_configs.delete_deployment_config_version(
+ workspace="workspace",
+ config="config",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**config:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_deployment_configs.get_latest_deployment_config(...) -> ModelDeploymentConfig
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get the latest version of a ModelDeploymentConfig.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_deployment_configs.get_latest_deployment_config(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_deployment_configs.update_deployment_config(...) -> ModelDeploymentConfig
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Update a ModelDeploymentConfig (creates a new immutable version).
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi, ModelDeploymentConfigModelSpec, ContainerExecutorConfig
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_deployment_configs.update_deployment_config(
+ workspace="workspace",
+ name="name",
+ engine="nim",
+ model_spec=ModelDeploymentConfigModelSpec(),
+ executor_config=ContainerExecutorConfig(
+ gpu=1,
+ ),
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**engine:** `Engine` — Inference engine selecting the compiler path (nim/vllm/generic)
+
+
+
+
+
+-
+
+**model_spec:** `ModelDeploymentConfigModelSpec` — What model to serve and how -- independent of the executor it runs on
+
+
+
+
+
+-
+
+**executor_config:** `ContainerExecutorConfig` — Compute + container settings for the executor the deployment runs on
+
+
+
+
+
+-
+
+**description:** `typing.Optional[str]` — Optional description of the deployment configuration
+
+
+
+
+
+-
+
+**model_entity_id:** `typing.Optional[str]` — Optional reference to the base model entity ID for this deployment
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_deployment_configs.delete_all_deployment_config_versions(...)
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Delete all versions of a ModelDeploymentConfig.
+
+This operation will fail with 409 Conflict if any ModelDeployments currently
+reference this config and are not in DELETED status. Delete or wait for
+dependent deployments to reach DELETED status before deleting the config.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_deployment_configs.delete_all_deployment_config_versions(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_deployment_configs.list_deployment_config_versions(...) -> typing.List[ModelDeploymentConfig]
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+List all versions of a ModelDeploymentConfig.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_deployment_configs.list_deployment_config_versions(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_deployments.list_deployments(...) -> ModelDeploymentsPage
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+List ModelDeployments for a specific workspace.
+
+By default, returns only the latest version of each deployment.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_deployments.list_deployments(
+ workspace="workspace",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number.
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Page size.
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[str]` — The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+
+
+
+
+-
+
+**all_versions:** `typing.Optional[bool]` — If true, return all versions of each deployment. If false (default), return only the latest version.
+
+
+
+
+
+-
+
+**filter:** `typing.Optional[ModelDeploymentFilter]` — Filter deployments by workspace, project, status, config, model_provider_id, name, status_message, created_at, and updated_at.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_deployments.create_deployment(...) -> ModelDeployment
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Create a new ModelDeployment (version 1).
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_deployments.create_deployment(
+ workspace="workspace",
+ name="llama-deploy-v1",
+ config="config",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str` — Name of the deployment. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+
+
+
+
+-
+
+**config:** `str` — Reference to the ModelDeploymentConfig name
+
+
+
+
+
+-
+
+**project:** `typing.Optional[str]` — The URN of the project associated with this deployment
+
+
+
+
+
+-
+
+**config_version:** `typing.Optional[int]` — Reference to a specific ModelDeploymentConfig version. If not specified, uses latest.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_deployments.get_deployment_version(...) -> ModelDeployment
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get a specific version of a ModelDeployment.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_deployments.get_deployment_version(
+ workspace="workspace",
+ deployment="deployment",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**deployment:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_deployments.delete_deployment_version(...) -> typing.Optional[typing.Any]
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Delete a specific version of a ModelDeployment.
+
+If the deployment is in any state other than DELETED, this will set its status to DELETING.
+The models controller will then:
+1. Delete the infrastructure (e.g., K8s NimService)
+2. Update the status to DELETED
+
+If the deployment is already in DELETED status, calling delete again will permanently
+remove it from the database.
+
+Returns:
+- 202 Accepted: Deployment version marked for deletion (status set to DELETING)
+- 204 No Content: Deployment version permanently removed from database (was already DELETED)
+- 404 Not Found: Deployment version doesn't exist
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_deployments.delete_deployment_version(
+ workspace="workspace",
+ deployment="deployment",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**deployment:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_deployments.get_latest_deployment(...) -> ModelDeployment
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get the latest version of a ModelDeployment.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_deployments.get_latest_deployment(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_deployments.update_deployment(...) -> ModelDeployment
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Update a ModelDeployment (creates a new immutable version).
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_deployments.update_deployment(
+ workspace="workspace",
+ name="name",
+ config="config",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**config:** `str` — Reference to the ModelDeploymentConfig name
+
+
+
+
+
+-
+
+**config_version:** `typing.Optional[int]` — Reference to a specific ModelDeploymentConfig version. If not specified, uses latest.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_deployments.delete_all_deployment_versions(...) -> typing.Optional[typing.Any]
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Delete all versions of a ModelDeployment.
+
+If the deployment is in any state other than DELETED, this will set its status to DELETING.
+The models controller will then:
+1. Delete the infrastructure (e.g., K8s NimService)
+2. Update the status to DELETED
+
+If the deployment is already in DELETED status, calling delete again will permanently
+remove it from the database.
+
+Returns:
+- 202 Accepted: Deployment marked for deletion (status set to DELETING)
+- 204 No Content: Deployment permanently removed from database (was already DELETED)
+- 404 Not Found: Deployment doesn't exist
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_deployments.delete_all_deployment_versions(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_deployments.get_deployment_models(...) -> typing.Dict[str, typing.Any]
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get Latest ModelDeployment's Model Entities from Entity Store.
+This provides the API contract that NIMs expect from Entity Store today, for pulling LoRAs,
+but enables us to enforce AuthZ boundaries.
+
+TODO: Implement model entity retrieval based on deployment config.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_deployments.get_deployment_models(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_deployments.update_deployment_status(...) -> ModelDeployment
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Update the status of a ModelDeployment (mutable operation).
+If version is not specified, updates the latest version.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_deployments.update_deployment_status(
+ workspace="workspace",
+ name="name",
+ status="UNKNOWN",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**status:** `ModelDeploymentStatus` — New status for the deployment
+
+
+
+
+
+-
+
+**version:** `typing.Optional[str]`
+
+
+
+
+
+-
+
+**status_message:** `typing.Optional[str]` — Detailed status message
+
+
+
+
+
+-
+
+**model_provider_id:** `typing.Optional[str]` — Optional reference to the auto-created ModelProvider workspace/name (format: workspace/name)
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_deployments.list_deployment_versions(...) -> typing.List[ModelDeployment]
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+List all versions of a ModelDeployment.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_deployments.list_deployment_versions(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.models.list_models(...) -> ModelEntitysPage
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+List Models endpoint with filtering, pagination, and sorting.
+
+Supports filter parameters for various criteria (including peft, custom fields),
+pagination (page, page_size), sorting, and workspace filtering via query parameter.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.models.list_models(
+ workspace="workspace",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number.
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Page size.
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[ModelEntitySortField]` — The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+
+
+
+
+-
+
+**verbose:** `typing.Optional[bool]` — Whether to include full spec details
+
+
+
+
+
+-
+
+**filter:** `typing.Optional[ModelEntityFilter]` — Filter models by name, project, workspace, base_model, adapters, finetuning_type, prompt, lora_enabled, description, created_at, and updated_at.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.models.create_model(...) -> ModelEntity
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Create a new model entity.
+
+This endpoint creates a new Model Entity in the Models service database.
+The Model Entity will be registered for use within the platform.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.models.create_model(
+ workspace="workspace",
+ name="llama-3.1-8b",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str` — Name of the model entity. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+
+
+
+
+-
+
+**project:** `typing.Optional[str]` — The URN of the project associated with this model entity
+
+
+
+
+
+-
+
+**description:** `typing.Optional[str]` — Optional description of the model
+
+
+
+
+
+-
+
+**spec:** `typing.Optional[ModelSpec]` — Detailed specification for the model - Automatically generated by the platform at creation when fileset provided.
+
+
+
+
+
+-
+
+**finetuning_type:** `typing.Optional[FinetuningType]` — Set for full weight finetuned models
+
+
+
+
+
+-
+
+**fileset:** `typing.Optional[str]` — A set of checkpoint files, configs, and other auxiliary info associated with this model - expected format {workspace}/{fileset_name}
+
+
+
+
+
+-
+
+**base_model:** `typing.Optional[str]` — Link to another model which is used as a base for the current model
+
+
+
+
+
+-
+
+**api_endpoint:** `typing.Optional[ApiEndpointData]` — Data about the inference endpoint for this model
+
+
+
+
+
+-
+
+**backend_format:** `typing.Optional[BackendFormat]` — Inference API wire format expected by the backend. If unset, inference routing treats the model as OPENAI_CHAT.
+
+
+
+
+
+-
+
+**prompt:** `typing.Optional[PromptData]` — Configuration for prompt engineering
+
+
+
+
+
+-
+
+**custom_fields:** `typing.Optional[typing.Dict[str, typing.Any]]` — Custom fields for additional metadata
+
+
+
+
+
+-
+
+**ownership:** `typing.Optional[typing.Dict[str, typing.Any]]` — Ownership information for the model
+
+
+
+
+
+-
+
+**model_providers:** `typing.Optional[typing.List[str]]` — List of ModelProvider workspace/name resource names that provide inference for this Model Entity
+
+
+
+
+
+-
+
+**trust_remote_code:** `typing.Optional[bool]`
+
+Whether to trust remote code for the checkpoint.
+ Some models without support in certain libraries such as Transformers require additional custom Python code to execute.
+ Due to security ramifications of running arbitrary code, this can only be set to true on one of the following conditions:
+ (1) the model's fileset's source is pre-approved in the platform config, or
+ (2) the user creating this model is an administrator.
+
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.models.create_model_adapter(...) -> Adapter
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Adds an Adapter to the Model
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.models.create_model_adapter(
+ workspace="workspace",
+ model_name="model_name",
+ name="lora-adapter-v1",
+ fileset="fileset",
+ finetuning_type="lora_merged",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**model_name:** `str`
+
+
+
+
+
+-
+
+**name:** `str` — Name of the adapter. Name must be unique in the workspace. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+
+
+
+
+-
+
+**fileset:** `str` — Location where adapter files are stored - expected format {workspace}/{fileset_name}
+
+
+
+
+
+-
+
+**finetuning_type:** `FinetuningType` — Type of finetuning (LORA, P_TUNING, etc.)
+
+
+
+
+
+-
+
+**description:** `typing.Optional[str]` — Optional description of the adapter
+
+
+
+
+
+-
+
+**enabled:** `typing.Optional[bool]` — Whether to make this adapter available for inference post training
+
+
+
+
+
+-
+
+**lora_config:** `typing.Optional[Lora]` — Lora configuration specifics
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.models.delete_model_adapter(...)
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Delete Adapter from Model entity.
+
+Permanently deletes an adapter from a model entity, if it was deployed, it will be cleaned up automatically.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.models.delete_model_adapter(
+ workspace="workspace",
+ model_name="model_name",
+ adapter="adapter",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**model_name:** `str`
+
+
+
+
+
+-
+
+**adapter:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.models.update_model_adapter(...) -> Adapter
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Update Adapter deployment or description.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.models.update_model_adapter(
+ workspace="workspace",
+ model_name="model_name",
+ adapter="adapter",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**model_name:** `str`
+
+
+
+
+
+-
+
+**adapter:** `str`
+
+
+
+
+
+-
+
+**request:** `UpdateAdapterRequest`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.models.get_model(...) -> ModelEntity
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get Model by Workspace and Name.
+
+Returns the details of a specific model entity identified by its workspace and name.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.models.get_model(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**verbose:** `typing.Optional[bool]` — Whether to include full spec details
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.models.delete_model(...)
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Delete Model entity.
+
+Permanently deletes a model entity from the platform.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.models.delete_model(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.models.update_model(...) -> ModelEntity
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Update Model metadata.
+
+Updates the metadata of an existing model entity. If the request body has an empty field,
+the old value is kept.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.models.update_model(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**verbose:** `typing.Optional[bool]` — Whether to include full spec details
+
+
+
+
+
+-
+
+**description:** `typing.Optional[str]` — Optional description of the model
+
+
+
+
+
+-
+
+**spec:** `typing.Optional[ModelSpec]` — Detailed specification for the model
+
+
+
+
+
+-
+
+**fileset:** `typing.Optional[str]` — A set of checkpoint files, configs, and other auxiliary info associated with this model - expected format {workspace}/{fileset_name}
+
+
+
+
+
+-
+
+**finetuning_type:** `typing.Optional[FinetuningType]` — Set for full weight finetuned models
+
+
+
+
+
+-
+
+**base_model:** `typing.Optional[str]` — Link to another model which is used as a base for the current model
+
+
+
+
+
+-
+
+**api_endpoint:** `typing.Optional[ApiEndpointData]` — Data about the inference endpoint for this model
+
+
+
+
+
+-
+
+**backend_format:** `typing.Optional[BackendFormat]` — Inference API wire format expected by the backend. If unset, inference routing treats the model as OPENAI_CHAT.
+
+
+
+
+
+-
+
+**prompt:** `typing.Optional[PromptData]` — Configuration for prompt engineering
+
+
+
+
+
+-
+
+**custom_fields:** `typing.Optional[typing.Dict[str, typing.Any]]` — Custom fields for additional metadata
+
+
+
+
+
+-
+
+**ownership:** `typing.Optional[typing.Dict[str, typing.Any]]` — Ownership information for the model
+
+
+
+
+
+-
+
+**model_providers:** `typing.Optional[typing.List[str]]` — List of ModelProvider workspace/name resource names that provide inference for this Model Entity
+
+
+
+
+
+-
+
+**trust_remote_code:** `typing.Optional[bool]`
+
+Whether to trust remote code for the checkpoint.
+ Some models without support in certain libraries such as Transformers require additional custom Python code to execute.
+ Due to security ramifications of running arbitrary code, this can only be set to true on one of the following conditions:
+ (1) the model's fileset's source is pre-approved in the platform config, or
+ (2) the user creating this model is an administrator.
+
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_providers.list_providers(...) -> ModelProvidersPage
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+List model providers for a specific workspace.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_providers.list_providers(
+ workspace="workspace",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number.
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Page size.
+
+
+
+
+
+-
+
+**sort:** `typing.Optional[ModelProviderSort]` — The field to sort by. To sort in decreasing order, use `-` in front of the field name.
+
+
+
+
+
+-
+
+**filter:** `typing.Optional[ModelProviderFilter]` — Filter model providers by workspace, project, status, model_deployment_id, name, description, host_url, created_at, and updated_at.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_providers.create_provider(...) -> ModelProvider
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Create a new model provider.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_providers.create_provider(
+ workspace="workspace",
+ name="my-nim-provider",
+ host_url="host_url",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str` — Name of the model provider. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+
+
+
+
+-
+
+**host_url:** `str` — The network endpoint URL for the model provider
+
+
+
+
+
+-
+
+**project:** `typing.Optional[str]` — The URN of the project associated with this model provider
+
+
+
+
+
+-
+
+**description:** `typing.Optional[str]` — Optional description of the model provider
+
+
+
+
+
+-
+
+**api_key_secret_name:** `typing.Optional[str]` — Reference to an API key secret stored in the Secrets service. Create the secret first via secrets API, then pass the secret name here.
+
+
+
+
+
+-
+
+**enabled_models:** `typing.Optional[typing.List[str]]` — Optional list of specific models to enable from this provider
+
+
+
+
+
+-
+
+**default_extra_body:** `typing.Optional[typing.Dict[str, typing.Any]]` — Default body parameters for inference requests. Can be overridden by user requests.
+
+
+
+
+
+-
+
+**default_extra_headers:** `typing.Optional[typing.Dict[str, str]]` — Default headers for inference requests. Can be overridden by user requests.
+
+
+
+
+
+-
+
+**required_extra_body:** `typing.Optional[typing.Dict[str, typing.Any]]` — Required body parameters for inference requests. Cannot be overridden by user requests.
+
+
+
+
+
+-
+
+**required_extra_headers:** `typing.Optional[typing.Dict[str, str]]` — Required headers for inference requests. Cannot be overridden by user requests.
+
+
+
+
+
+-
+
+**model_deployment_id:** `typing.Optional[str]` — Optional reference to the ModelDeployment ID if this provider is being auto-created for a deployment
+
+
+
+
+
+-
+
+**status:** `typing.Optional[ModelProviderStatus]` — Status of the model provider
+
+
+
+
+
+-
+
+**status_message:** `typing.Optional[str]` — Status message
+
+
+
+
+
+-
+
+**auth_header_format:** `typing.Optional[str]` — Jinja2 template string controlling how the API key secret is sent to the upstream. Must contain exactly one variable named `auth_secret`, which is substituted with the resolved secret value at request time. Example: `'X-Api-Key: {{ auth_secret }}'`. If not set, defaults to `'Authorization: Bearer {{ auth_secret }}'`.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_providers.get_provider(...) -> ModelProvider
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Get a model provider by workspace and name.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_providers.get_provider(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_providers.upsert_provider(...) -> ModelProvider
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Create or update a model provider.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_providers.upsert_provider(
+ workspace="workspace",
+ name="name",
+ host_url="host_url",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**host_url:** `str` — The network endpoint URL for the model provider
+
+
+
+
+
+-
+
+**project:** `typing.Optional[str]` — The URN of the project associated with this model provider
+
+
+
+
+
+-
+
+**description:** `typing.Optional[str]` — Optional description of the model provider
+
+
+
+
+
+-
+
+**api_key_secret_name:** `typing.Optional[str]` — Reference to an API key secret stored in the Secrets service. Create the secret first via secrets API, then pass the secret name here.
+
+
+
+
+
+-
+
+**enabled_models:** `typing.Optional[typing.List[str]]` — Optional list of specific models to enable from this provider
+
+
+
+
+
+-
+
+**default_extra_body:** `typing.Optional[typing.Dict[str, typing.Any]]` — Default body parameters for inference requests. Can be overridden by user requests.
+
+
+
+
+
+-
+
+**default_extra_headers:** `typing.Optional[typing.Dict[str, str]]` — Default headers for inference requests. Can be overridden by user requests.
+
+
+
+
+
+-
+
+**required_extra_body:** `typing.Optional[typing.Dict[str, typing.Any]]` — Required body parameters for inference requests. Cannot be overridden by user requests.
+
+
+
+
+
+-
+
+**required_extra_headers:** `typing.Optional[typing.Dict[str, str]]` — Required headers for inference requests. Cannot be overridden by user requests.
+
+
+
+
+
+-
+
+**model_deployment_id:** `typing.Optional[str]` — Optional reference to the ModelDeployment ID if this provider is associated with a deployment
+
+
+
+
+
+-
+
+**status:** `typing.Optional[ModelProviderStatus]` — Status of the model provider
+
+
+
+
+
+-
+
+**status_message:** `typing.Optional[str]` — Status message
+
+
+
+
+
+-
+
+**auth_header_format:** `typing.Optional[str]` — Jinja2 template string controlling how the API key secret is sent to the upstream. Must contain exactly one variable named `auth_secret`, which is substituted with the resolved secret value at request time. Example: `'X-Api-Key: {{ auth_secret }}'`. If not set, defaults to `'Authorization: Bearer {{ auth_secret }}'`.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_providers.delete_provider(...)
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Delete a model provider by workspace and name.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_providers.delete_provider(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.model_providers.update_provider_status(...) -> ModelProvider
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Update status-related fields of a model provider.
+
+This endpoint supports partial updates for fields managed by Models Controller:
+- model_deployment_id
+- served_models
+- status
+- status_message
+
+If status is provided without status_message, status_message will be set to empty string.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.model_providers.update_provider_status(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**model_deployment_id:** `typing.Optional[str]` — Reference to the ModelDeployment ID if this provider is associated with a deployment
+
+
+
+
+
+-
+
+**served_models:** `typing.Optional[typing.List[ServedModelMapping]]` — List of models served by this provider with routing information for IGW
+
+
+
+
+
+-
+
+**status:** `typing.Optional[ModelProviderStatus]` — Status of the model provider
+
+
+
+
+
+-
+
+**status_message:** `typing.Optional[str]` — Status message. If status is provided without status_message, defaults to empty string.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.secrets_admin.admin_rotate_encryption_keys() -> PlatformSecretAdminRotationResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Rotate encryption keys for all platform secrets.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.secrets_admin.admin_rotate_encryption_keys()
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.secrets.list_secrets(...) -> PlatformSecretResponsesPage
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+List available secrets
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.secrets.list_secrets(
+ workspace="workspace",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**page:** `typing.Optional[int]` — Page number.
+
+
+
+
+
+-
+
+**page_size:** `typing.Optional[int]` — Page size.
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.secrets.create_secret(...) -> PlatformSecretResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Create a new secret.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.secrets.create_secret(
+ workspace="workspace",
+ name="hf-token",
+ value="value",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str` — The name of the secret to create. Allowed characters: letters (a-z, A-Z), digits (0-9), underscores, hyphens, and dots.
+
+
+
+
+
+-
+
+**value:** `str` — The payload of the secret
+
+
+
+
+
+-
+
+**description:** `typing.Optional[str]` — An optional description of the secret
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.secrets.get_secret(...) -> PlatformSecretResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Retrieve a secret by its name.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.secrets.get_secret(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.secrets.delete_secret(...)
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Delete a secret.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.secrets.delete_secret(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.secrets.update_secret(...) -> PlatformSecretResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Update a secret's metadata.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.secrets.update_secret(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**description:** `typing.Optional[str]` — An optional description of the secret
+
+
+
+
+
+-
+
+**value:** `typing.Optional[str]` — The new secret value
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+client.secrets.access_secret(...) -> PlatformSecretAccessResponse
+
+-
+
+#### 📝 Description
+
+
+-
+
+
+-
+
+Access the value of a secret.
+
+
+
+
+
+#### 🔌 Usage
+
+
+-
+
+
+-
+
+```python
+from nvidia import NvidiaApi
+
+client = NvidiaApi(
+ base_url="https://yourhost.com/path/to/api",
+)
+
+client.secrets.access_secret(
+ workspace="workspace",
+ name="name",
+)
+
+```
+
+
+
+
+
+#### ⚙️ Parameters
+
+
+-
+
+
+-
+
+**workspace:** `str`
+
+
+
+
+
+-
+
+**name:** `str`
+
+
+
+
+
+-
+
+**request_options:** `typing.Optional[RequestOptions]` — Request-specific configuration.
+
+
+
+
+
+
+
+
+
+