fix: sync sdk with orchestration spec

packages/gen/gen_ai_hub/orchestration_v2/__init__.py

@@ -11,6 +11,7 @@
 
     # config
     "ModuleConfig", "OrchestrationConfig", "OrchestrationConfigReference",
+    "PartialPromptTemplatingModuleConfig", "PartialModuleConfigs", "PartialOrchestrationConfig",
     "CompletionRequestConfigurationReferenceByIdConfigRef",
     "CompletionRequestConfigurationReferenceByNameScenarioVersionConfigRef",
 
@@ -22,8 +23,9 @@
     "InputFiltering", "OutputFiltering", "FilteringModuleConfig",
 
     # data_masking
-    "DataMaskingProviderName", "MaskingMethod", "ProfileEntity", "DPIMethodConstant", "DPIMethodFabricatedData",
-    "DPICustomEntity", "DPIStandardEntity", "MaskGroundingInput", "MaskingProviderConfig", "MaskingModuleConfig",
+    "DataMaskingProviderName", "MaskingMethod", "MaskFileInputMethod", "ProfileEntity", "DPIMethodConstant",
+    "DPIMethodFabricatedData", "DPICustomEntity", "DPIStandardEntity", "MaskGroundingInput", "MaskingProviderConfig",
+    "MaskingModuleConfig",
 
     # document_grounding
     "GroundingType", "DataRepositoryType", "DocumentGroundingFilter", "DocumentGroundingPlaceholders",
@@ -33,7 +35,8 @@
     # embeddings
     "EmbeddingsEncodingFormat", "EmbeddingsInputType", "EmbeddingsModelParams", "EmbeddingsModelDetails",
     "EmbeddingsModelConfig", "EmbeddingsModuleConfigs", "EmbeddingsOrchestrationConfig", "EmbeddingsInput",
-    "EmbeddingsUsage", "EmbeddingResult", "EmbeddingsResponse", "EmbeddingsPostResponse", "EmbeddingsRequest",
+    "EmbeddingsUsage", "EmbeddingResult", "EmbeddingsResponse",
+    "EmbeddingsPostResponse", "EmbeddingsRequest",
 
     # llama_guard_3_filter
     "LlamaGuard38bFilter",
@@ -47,6 +50,7 @@
 
     # multimodal_items
     "ImageDetailLevel", "TextPart", "ImageUrl", "ImagePart", "ContentPart", "ImageItem",
+    "FileContent", "FilePart", "FileItem",
 
     # response
     "CacheCreationTokenDetails", "PromptTokensDetails", "CompletionTokensDetails", "TokenUsage", "GenericModuleResult", "TopLogprob",

packages/gen/gen_ai_hub/orchestration_v2/models/__init__.py

@@ -1,26 +1,29 @@
 from .azure_content_filter import AzureContentSafetyInput, AzureContentSafetyOutput, AzureContentFilter, AzureThreshold
 from .cache_control import CacheControl
 from .config import (ModuleConfig, OrchestrationConfig, OrchestrationConfigReference,
+                     PartialPromptTemplatingModuleConfig, PartialModuleConfigs, PartialOrchestrationConfig,
                      CompletionRequestConfigurationReferenceByIdConfigRef,
                      CompletionRequestConfigurationReferenceByNameScenarioVersionConfigRef)
 from .content_filter import (ContentFilterProvider, ContentFilter, LlamaGuard38bFilterConfig,
                              AzureContentSafetyInputFilterConfig, AzureContentSafetyOutputFilterConfig,
                              FilteringStreamOptions)
 from .content_filtering import InputFiltering, OutputFiltering, FilteringModuleConfig
-from .data_masking import (DataMaskingProviderName, MaskingMethod, ProfileEntity, DPIMethodConstant, 
-                          DPIMethodFabricatedData, DPICustomEntity, DPIStandardEntity, MaskGroundingInput, 
-                          MaskingProviderConfig, MaskingModuleConfig)
+from .data_masking import (DataMaskingProviderName, MaskingMethod, MaskFileInputMethod, ProfileEntity,
+                          DPIMethodConstant, DPIMethodFabricatedData, DPICustomEntity, DPIStandardEntity,
+                          MaskGroundingInput, MaskingProviderConfig, MaskingModuleConfig)
 from .document_grounding import (GroundingType, DataRepositoryType, DocumentGroundingFilter, 
                                 DocumentGroundingPlaceholders, DocumentGroundingConfig, GroundingModuleConfig, 
                                 KeyValueListPair, DocumentMetadataKeyValueListPairs, GroundingSearchConfig)
 from .embeddings import (EmbeddingsEncodingFormat, EmbeddingsInputType, EmbeddingsModelParams, EmbeddingsModelDetails,
                         EmbeddingsModelConfig, EmbeddingsModuleConfigs, EmbeddingsOrchestrationConfig, EmbeddingsInput,
-                        EmbeddingsUsage, EmbeddingResult, EmbeddingsResponse, EmbeddingsPostResponse, EmbeddingsRequest)
+                        EmbeddingsUsage, EmbeddingResult, EmbeddingsResponse,
+                        EmbeddingsPostResponse, EmbeddingsRequest)
 from .llama_guard_3_filter import LlamaGuard38bFilter
 from .llm_model_details import LLMModelDetails
 from .message import (SystemMessage, UserMessage, AssistantMessage, ToolChatMessage, DeveloperChatMessage, ChatMessage,
                      ResponseChatMessage, FunctionCall, MessageToolCall, ReasoningBlock)
-from .multimodal_items import ImageDetailLevel, TextPart, ImageUrl, ImagePart, ContentPart, ImageItem
+from .multimodal_items import (ImageDetailLevel, TextPart, ImageUrl, ImagePart, ContentPart, ImageItem,
+                               FileContent, FilePart, FileItem)
 from .response import (CacheCreationTokenDetails, PromptTokensDetails, CompletionTokensDetails, TokenUsage, GenericModuleResult, TopLogprob,
                       ChatCompletionTokenLogprob, ChoiceLogprobs, LLMChoice, StreamFunctionObject, StreamToolCall,
                       StreamDelta, StreamLLMChoice, Citation, LLMModuleResult, StreamLLMModuleResult, ModuleResults,
@@ -45,6 +48,7 @@
 
     # config
     "ModuleConfig", "OrchestrationConfig", "OrchestrationConfigReference",
+    "PartialPromptTemplatingModuleConfig", "PartialModuleConfigs", "PartialOrchestrationConfig",
     "CompletionRequestConfigurationReferenceByIdConfigRef",
     "CompletionRequestConfigurationReferenceByNameScenarioVersionConfigRef",
 
@@ -56,8 +60,9 @@
     "InputFiltering", "OutputFiltering", "FilteringModuleConfig",
 
     # data_masking
-    "DataMaskingProviderName", "MaskingMethod", "ProfileEntity", "DPIMethodConstant", "DPIMethodFabricatedData",
-    "DPICustomEntity", "DPIStandardEntity", "MaskGroundingInput", "MaskingProviderConfig", "MaskingModuleConfig",
+    "DataMaskingProviderName", "MaskingMethod", "MaskFileInputMethod", "ProfileEntity", "DPIMethodConstant",
+    "DPIMethodFabricatedData", "DPICustomEntity", "DPIStandardEntity", "MaskGroundingInput", "MaskingProviderConfig",
+    "MaskingModuleConfig",
 
     # document_grounding
     "GroundingType", "DataRepositoryType", "DocumentGroundingFilter", "DocumentGroundingPlaceholders",
@@ -67,7 +72,8 @@
     # embeddings
     "EmbeddingsEncodingFormat", "EmbeddingsInputType", "EmbeddingsModelParams", "EmbeddingsModelDetails",
     "EmbeddingsModelConfig", "EmbeddingsModuleConfigs", "EmbeddingsOrchestrationConfig", "EmbeddingsInput",
-    "EmbeddingsUsage", "EmbeddingResult", "EmbeddingsResponse", "EmbeddingsPostResponse", "EmbeddingsRequest",
+    "EmbeddingsUsage", "EmbeddingResult", "EmbeddingsResponse",
+    "EmbeddingsPostResponse", "EmbeddingsRequest",
 
     # llama_guard_3_filter
     "LlamaGuard38bFilter",
@@ -81,14 +87,15 @@
 
     # multimodal_items
     "ImageDetailLevel", "TextPart", "ImageUrl", "ImagePart", "ContentPart", "ImageItem",
+    "FileContent", "FilePart", "FileItem",
 
     # response
     "CacheCreationTokenDetails", "PromptTokensDetails", "CompletionTokensDetails", "TokenUsage", "GenericModuleResult", "TopLogprob",
     "ChatCompletionTokenLogprob", "ChoiceLogprobs", "LLMChoice", "StreamFunctionObject", "StreamToolCall",
     "StreamDelta", "StreamLLMChoice", "Citation", "LLMModuleResult", "StreamLLMModuleResult", "ModuleResults",
     "StreamModuleResults", "SAPAPIError", "SAPAPIErrorStreaming", "CompletionPostResponse",
     "StreamCompletionPostResponse", "ErrorResponse", "ErrorResponseStreaming", "OrchestrationResponseWithRetries",
-    
+
     # response_format
     "ResponseFormatText", "ResponseFormatJsonObject", "ResponseFormatJsonSchema", "JSONResponseSchema",
 

packages/gen/gen_ai_hub/orchestration_v2/models/config.py

@@ -10,8 +10,10 @@
 from gen_ai_hub.orchestration_v2.models.content_filtering import FilteringModuleConfig
 from gen_ai_hub.orchestration_v2.models.data_masking import MaskingModuleConfig
 from gen_ai_hub.orchestration_v2.models.document_grounding import GroundingModuleConfig
+from gen_ai_hub.orchestration_v2.models.llm_model_details import LLMModelDetails
 from gen_ai_hub.orchestration_v2.models.streaming import GlobalStreamOptions
-from gen_ai_hub.orchestration_v2.models.template import PromptTemplatingModuleConfig
+from gen_ai_hub.orchestration_v2.models.template import PromptTemplatingModuleConfig, Template
+from gen_ai_hub.orchestration_v2.models.template_ref import TemplateRef
 from gen_ai_hub.orchestration_v2.models.translation import TranslationModuleConfig
 
 
@@ -59,6 +61,62 @@ class OrchestrationConfig(BaseModel):
     stream: Optional[GlobalStreamOptions] = None
 
 
+class PartialPromptTemplatingModuleConfig(BaseModel):
+    """
+    Partial prompt templating configuration for use with config_ref overrides.
+
+    Every field is optional so that only the parts to override need to be provided;
+    the referenced configuration supplies the rest.
+
+    Args:
+        prompt: Optional prompt template override (inline Template or TemplateRef).
+
+        model: Optional model override. Unlike the non-partial config, model is optional
+            here so that only the prompt can be overridden without repeating the model config.
+    """
+    prompt: Optional[Template | TemplateRef] = None
+    model: Optional[LLMModelDetails] = None
+
+
+class PartialModuleConfigs(BaseModel):
+    """
+    Partial module configuration for use with config_ref overrides.
+
+    Every module is optional; only the modules provided here override the referenced config.
+
+    Args:
+        prompt_templating: Optional partial prompt templating override.
+
+        filtering: Optional filtering module override.
+
+        masking: Optional masking module override.
+
+        grounding: Optional grounding module override.
+
+        translation: Optional translation module override.
+    """
+    prompt_templating: Optional[PartialPromptTemplatingModuleConfig] = None
+    filtering: Optional[FilteringModuleConfig] = None
+    masking: Optional[MaskingModuleConfig] = None
+    grounding: Optional[GroundingModuleConfig] = None
+    translation: Optional[TranslationModuleConfig] = None
+
+
+class PartialOrchestrationConfig(BaseModel):
+    """
+    Partial orchestration configuration used to override a referenced config (config_ref).
+
+    The referenced configuration is the base; the fields provided here are merged on top.
+
+    Args:
+        modules: Optional partial module configurations to override.
+
+        stream: Optional streaming configuration to override.
+    """
+    modules: Optional[PartialModuleConfigs] = None
+    stream: Optional[GlobalStreamOptions] = None
+
+
 class CompletionRequestConfigurationReferenceByIdConfigRef(BaseModel):
     """Represents a reference to an orchestration config identified by a unique ID.
 
@@ -86,5 +144,6 @@ class CompletionRequestConfigurationReferenceByNameScenarioVersionConfigRef(Base
                                 CompletionRequestConfigurationReferenceByNameScenarioVersionConfigRef)
 
 __all__ = ["ModuleConfig", "OrchestrationConfig", "OrchestrationConfigReference",
+           "PartialPromptTemplatingModuleConfig", "PartialModuleConfigs", "PartialOrchestrationConfig",
            "CompletionRequestConfigurationReferenceByIdConfigRef",
            "CompletionRequestConfigurationReferenceByNameScenarioVersionConfigRef"]

packages/gen/gen_ai_hub/orchestration_v2/models/data_masking.py

@@ -166,6 +166,21 @@ class MaskGroundingInput(BaseModel):
     """
     enabled: bool = False
 
+class MaskFileInputMethod(str, Enum):
+    """
+    Enumerates the supported masking methods for file inputs.
+
+    This enum defines how file inputs are handled by the data masking module. It is required if file inputs
+    are provided.
+
+    Values:
+        ANONYMIZATION: Irreversibly replaces sensitive data in file inputs with placeholders.
+
+        SKIP: Skips masking of file inputs, leaving them unchanged.
+    """
+    ANONYMIZATION = "anonymization"
+    SKIP = "skip"
+
 
 class MaskingProviderConfig(BaseModel):
     """
@@ -183,12 +198,15 @@ class MaskingProviderConfig(BaseModel):
         allowlist: A list of strings that should not be masked.
 
         mask_grounding_input: A flag indicating whether to mask input to the grounding module.
+
+        mask_file_input_method: The method of masking to apply to file inputs. Required if file inputs are provided.
     """
     type_: DataMaskingProviderName = Field(default=DataMaskingProviderName.SAP_DATA_PRIVACY_INTEGRATION, alias="type")
     method: MaskingMethod
     entities: List[Union[DPIStandardEntity, DPICustomEntity]]
     allowlist: Optional[List[str]] = None
     mask_grounding_input: Optional[MaskGroundingInput] = None
+    mask_file_input_method: Optional[MaskFileInputMethod] = None
 
 class MaskingModuleConfig(BaseModel):
     """
@@ -217,5 +235,6 @@ def enforce_exactly_one_provider_list(self):
 
         return self
 
-__all__ = ["DataMaskingProviderName", "MaskingMethod", "ProfileEntity", "DPIMethodConstant", "DPIMethodFabricatedData",
-           "DPICustomEntity", "DPIStandardEntity", "MaskGroundingInput", "MaskingProviderConfig", "MaskingModuleConfig"]
+__all__ = ["DataMaskingProviderName", "MaskingMethod", "MaskFileInputMethod", "ProfileEntity", "DPIMethodConstant",
+           "DPIMethodFabricatedData", "DPICustomEntity", "DPIStandardEntity", "MaskGroundingInput",
+           "MaskingProviderConfig", "MaskingModuleConfig"]

packages/gen/gen_ai_hub/orchestration_v2/models/embeddings.py

@@ -16,13 +16,19 @@ class EmbeddingsEncodingFormat(str, Enum):
     Encoding format for the embeddings output.
 
     Values:
-        FLOAT: Returns embeddings as an array of floats.
+        FLOAT: Returns embeddings as an array of floating point numbers.
         BASE64: Returns embeddings as a base64 encoded string.
         BINARY: Returns embeddings in binary format.
+        INT8: Returns embeddings as an array of 8-bit integers.
+        UINT8: Returns embeddings as an array of unsigned 8-bit integers.
+        UBINARY: Returns embeddings in unsigned binary format.
     """
     FLOAT = "float"
     BASE64 = "base64"
     BINARY = "binary"
+    INT8 = "int8"
+    UINT8 = "uint8"
+    UBINARY = "ubinary"
 
 
 class EmbeddingsInputType(str, Enum):
@@ -35,10 +41,20 @@ class EmbeddingsInputType(str, Enum):
         TEXT: General purpose text (default).
         DOCUMENT: Content to be searched/retrieved.
         QUERY: Short search queries.
+        SEARCH_DOCUMENT: Content stored in a search index to be retrieved.
+        SEARCH_QUERY: Query run against a search index.
+        CLASSIFICATION: Text used in a classification task.
+        CLUSTERING: Text used in a clustering task.
+        IMAGE: Image input.
     """
     TEXT = "text"
     DOCUMENT = "document"
     QUERY = "query"
+    SEARCH_DOCUMENT = "search_document"
+    SEARCH_QUERY = "search_query"
+    CLASSIFICATION = "classification"
+    CLUSTERING = "clustering"
+    IMAGE = "image"
 
 
 class EmbeddingsModelParams(BaseModel):
@@ -47,7 +63,7 @@ class EmbeddingsModelParams(BaseModel):
 
     Args:
         dimensions: The number of dimensions for the output embeddings.
-        encoding_format: The format for the embeddings output (float, base64, or binary).
+        encoding_format: The format to return the embeddings in.
         normalize: Whether to normalize the embeddings.
     """
     dimensions: Optional[int] = None
@@ -69,7 +85,7 @@ class EmbeddingsModelDetails(BaseModel):
     name: str
     version: Optional[str] = "latest"
     params: Optional[EmbeddingsModelParams] = None
-    timeout: Optional[int] = Field(default=600, ge=1, le=600)
+    timeout: Optional[int] = Field(default=600, ge=1, le=1200)
     max_retries: Optional[int] = Field(default=2, ge=0, le=5)
 
 
@@ -135,7 +151,7 @@ class EmbeddingResult(ResponseBaseModel):
 
     Args:
         object: The object type, always "embedding".
-        embedding: The embedding vector (array of floats) or base64 string.
+        embedding: The embedding vector (array of floats) or a base64 string.
         index: The index of this embedding in the list.
     """
     object: str = "embedding"
@@ -186,4 +202,5 @@ class EmbeddingsRequest(BaseModel):
 
 __all__ = ["EmbeddingsEncodingFormat", "EmbeddingsInputType", "EmbeddingsModelParams", "EmbeddingsModelDetails",
            "EmbeddingsModelConfig", "EmbeddingsModuleConfigs", "EmbeddingsOrchestrationConfig", "EmbeddingsInput",
-           "EmbeddingsUsage", "EmbeddingResult", "EmbeddingsResponse", "EmbeddingsPostResponse", "EmbeddingsRequest"]
+           "EmbeddingsUsage", "EmbeddingResult", "EmbeddingsResponse",
+           "EmbeddingsPostResponse", "EmbeddingsRequest"]

packages/gen/gen_ai_hub/orchestration_v2/models/llm_model_details.py

@@ -24,11 +24,18 @@ class LLMModelDetails(BaseModel):
         timeout: Timeout for the LLM request in seconds. This parameter is currently ignored for Vertex AI models.
 
         max_retries: Maximum number of retries for the LLM request. This parameter is currently ignored for Vertex AI models.
+
+        drop_unsupported_params: Controls behaviour when Chat Completions parameters not supported by the
+            Responses API are present in the request. Only applies to models that use the Responses API.
+            When True: unsupported params are silently dropped and reported in the
+            X-Orchestration-Dropped-Model-Params response header. When False: any unsupported parameter
+            causes a 400 error with an actionable message. Defaults to True.
     """
     name: str
     version: Optional[str] = "latest"
     params: Optional[Dict] = None
-    timeout: Optional[int] = Field(default=600, ge=1, le=600)
+    timeout: Optional[int] = Field(default=600, ge=1, le=1200)
     max_retries: Optional[int] = Field(default=2, ge=0, le=5)
+    drop_unsupported_params: Optional[bool] = True
 
 __all__ = ["LLMModelDetails"]

packages/gen/gen_ai_hub/orchestration_v2/models/message.py

@@ -11,7 +11,9 @@
 from pydantic import field_validator, ValidationError
 
 from gen_ai_hub.orchestration_v2.models.base import ABCBaseModel as BaseModel, ResponseBaseModel
-from gen_ai_hub.orchestration_v2.models.multimodal_items import ContentPart, ImageItem, TextPart, ImageUrl, ImagePart
+from gen_ai_hub.orchestration_v2.models.multimodal_items import (
+    ContentPart, ImageItem, TextPart, ImageUrl, ImagePart, FileItem, FilePart, FileContent
+)
 
 
 class FunctionCall(ResponseBaseModel):
@@ -122,10 +124,10 @@ class UserMessage(BaseModel):
     Args:
         role: The role of the entity sending the message.
 
-        content: The message content, which may be plain text or a sequence of text and images.
+        content: The message content, which may be plain text or a sequence of text, images, or files.
     """
     role: Role = Role.USER
-    content: Union[str, ContentPart, List[Union[str, ContentPart, ImageItem, ImagePart]]]
+    content: Union[str, ContentPart, List[Union[str, ContentPart, ImageItem, ImagePart, FileItem, FilePart]]]
 
     @field_validator("content", mode="before")
     def content_validation(cls, content):  # pylint: disable=no-self-argument
@@ -145,8 +147,10 @@ def content_validation(cls, content):  # pylint: disable=no-self-argument
                     mapped_content.append(TextPart(text=item))
                 elif isinstance(item, ImageItem):
                     mapped_content.append(ImagePart(image_url=ImageUrl(url=item.url, detail=item.detail)))
+                elif isinstance(item, FileItem):
+                    mapped_content.append(FilePart(file=FileContent(file_data=item.file_data, filename=item.filename)))
                 else:
-                    raise ValidationError("User message content list must contain only str or ImageItem")
+                    raise ValidationError("User message content list must contain only str, ImageItem or FileItem")
         return mapped_content