Generative Ai¶
OCI Generative AI is a fully managed service that provides a set of state-of-the-art, customizable large language models (LLMs) that cover a wide range of use cases for text generation, summarization, and text embeddings. |
|
This class provides a wrapper around |
Models¶
new artifact details. |
|
Results of an allocatable GPU search for launching dedicated AI clusters. |
|
Information about allocatable GPUs that can be used to launch dedicated AI clusters. |
|
Hostnames and ports allowed by a sandbox network policy. |
|
ApiKeys are resources used to access GenAI models. |
|
Results of a ApiKey search. |
|
The ApiKey item. |
|
Summary of the API key. |
|
Container/artifact configuration for the deployment. |
|
Describes where and how a model is deployed |
|
The details to move an APIKey to another compartment. |
|
|
The details to move a dedicated AI cluster to another compartment. |
The details to move an endpoint to another compartment. |
|
|
The details required to change a private endpoint compartment. |
|
The details to move a GenerativeAiProject to another compartment. |
|
The details to move a hosted application to another compartment. |
|
The details to move a hosted application storage to another compartment. |
|
The details to move an imported model to another compartment. |
The details to move a custom model to another compartment. |
|
|
The details to move a routing profile to another compartment. |
|
The details to move a SemanticStore to another compartment. |
The chat model metrics of the fine-tuning process. |
|
The compute capacity reservation in customer tenancy. |
|
Configuration for condensing conversation content. |
|
Datasource configuration for the connector. |
|
The configuration details, whether to add the content moderation feature to the model. |
|
Holds configuration related to conversation retention |
|
The data to create an API key. |
|
Artifact configuration input for the deployment. |
|
|
Defines the OCI Database Tools Connection data source that the semantic model connects to. |
Defines the data source that the semantic model connects to. |
|
The data to create a dedicated AI cluster. |
|
The data to create an endpoint. |
|
|
The details required to create a Generative AI private endpoint. |
The data to create a GenerativeAiProject. |
|
Common creation properties shared by HostedApplication and HostedApplicationIam. |
|
The details required to create a hosted application. |
|
The details required to create a hosted application. |
|
|
The data to create a hosted application storage. |
The data to create a hosted deployment. |
|
The data to import a model. |
|
The data to create a custom model. |
|
The data to create a routing profile. |
|
Create Sandbox Warm Pool Details. |
|
|
Array of database schemas or database objects included in the enrichment pipeline for data sources connected via an OCI Database Tools connection. |
Array of database schemas or other database objects to include in enrichment pipeline. |
|
The data to create a SemanticStore. |
|
CreateSingleDockerArtifactDetails model. |
|
The data to create a VectorStoreConnector. |
|
|
The data to create a VectorStoreConnectorFileSync. |
|
Outbound networking configuration that uses a customer subnet for sandboxes in this project. |
Uses a customer-selected generative AI model for SemanticStore enrichment. |
|
|
Defines the OCI Database Tools Connection data source that the semantic model connects to. |
Defines the data source that the semantic model connects to. |
|
DatabaseToolsConnection model. |
|
The dataset used to fine-tune the model. |
|
Dedicated AI clusters are compute resources that you can use for fine-tuning custom models or for hosting endpoints for custom models. |
|
The total capacity for a dedicated AI cluster. |
|
Results of a dedicate AI cluster search. |
|
The capacity of a hosting type dedicated AI cluster. |
|
Summary information about a dedicated AI cluster. |
|
Uses the service default generative AI model for SemanticStore enrichment. |
|
Configuration for generating embeddings from extracted information. |
|
To host a custom model for inference, create an endpoint for that model on a dedicated AI cluster of type HOSTING. |
|
Results of an endpoint search. |
|
Summary information for an endpoint resource. |
|
The environment variables for the Hosted Application |
|
Configuration for information extraction from conversation content. |
|
A Log object that gives the ingestion status of a File from a datasource read by a VectorStoreConnector |
|
Results of a VectorStoreConnector Ingestion Log search. |
|
Synchronization Statistics for a VectorStore File Sync operation or for a VectorStore Connector |
|
Details about fine-tuning a custom model. |
|
LLM selection with specific Gen AI model. |
|
Generative AI private endpoint. |
|
|
Collection of GenerativeAiPrivateEndpointSummary |
List of Generative AI private endpoints. |
|
A GenerativeAiProject is a logical container that stores conversation, file and containers. |
|
Results of a GenerativeAiProject search. |
|
Summary information for a GenerativeAiProject. |
|
GPU allocation details for an underlying capacity allocation. |
|
Hosted Application, defines shared configurations that apply across multiple deployments of the Agent or MCP application. |
|
Common properties shared by HostedApplication and HostedApplicationIam. |
|
Results of a dedicate hosted application search. |
|
Hosted Application IAM, defines shared configurations that apply across multiple deployments of the Agent or MCP application. |
|
Represents managed storage for an application. |
|
Results of a dedicate hosted application search. |
|
Summary information about a hosted application storage. |
|
Summary information about a hosted application. |
|
Hosted deployment is designed to support the full spectrum of agent use cases from lightweight, employee-facing assistants and internal workflow automation, to enterprise-grade, large-scale customer-facing workloads. |
|
Results of a dedicate hosted deployment search. |
|
Summary information about a hosted deployment. |
|
Configuration for importing a model from Hugging Face. |
|
Oracle Identity Cloud Service (IDCS) configuration used when inboundAuthConfigType is set to IDCS_AUTH_CONFIG or IDCS_SESSION_AUTH_CONFIG. |
|
Represents a model imported into the system based on an external data source, such as Hugging Face or Object Storage. |
|
Represents the result of a list operation for imported models. |
|
Summary of the importedModel. |
|
The inbound authentication configuration for the Hosted Application. |
|
Inbound Networking configuration. |
|
The data to create/renew an API key item. |
|
LLM selection configuration. |
|
Configuration settings for long-term memory behavior. |
|
The Lora training method hyperparameters. |
|
|
Outbound networking configuration that uses Oracle-managed networking for sandboxes in this project. |
You can create a custom model by using your dataset to fine-tune an out-of-the-box text generation base model. |
|
Results of a model search. |
|
Base model configuration shared across GenAI Project memory. |
|
Defines the source location and method used to import the model. |
|
Metadata describing a specific model |
|
A paginated collection of model discovery results. |
|
Model metrics during the creation of a new model. |
|
A supported input-to-output modality transformation for a model. |
|
The model routing policy for the routing profile. |
|
Summary of the model. |
|
Networking configuration. |
|
OCI Object storage configuration details. |
|
The dataset is stored in an OCI Object Storage bucket. |
|
Details about the object storage location. |
|
The OCI Object Storage namespace and bucket details of the datasource |
|
Outbound Networking configuration. |
|
Defines a configurable parameter supported by a model. |
|
The configuration details for personally identifiable information (PII) detection, in prompts and responses. |
|
A bounded snapshot of the previous state of a routing profile. |
|
The configuration details for prompt injection (PI) detection. |
|
Specifies a refresh schedule. |
|
Defines the refresh schedule by specifying the interval between each refresh. |
|
Allows user to opt out of automated synchronization schedule. |
|
Only trigger the enrichment at the creation time. |
|
The cross-region routing policy for the routing profile. |
|
The data to renew an API key item. |
|
A routing profile is a reusable set of policies that controls how Smart Router selects a target region and model for inference traffic. |
|
Results of a routing profile search. |
|
Summary information for a routing profile resource. |
|
Network policy that allows all network access for sandboxes in this project. |
|
Network policy that allows access only to specified domains for sandboxes in this project. |
|
Sandbox configuration under a project. |
|
The HTTP request header and OCI Vault secret used by a credential injection rule. |
|
Base type for a credential injection rule. |
|
A named Project-local credential profile selected explicitly when a Sandbox is created. |
|
Network policy that denies all network access for sandboxes in this project. |
|
|
An HTTP_HEADER rule that binds one exact HTTPS host on port 443 to exactly one injected HTTP request header and one OCI Vault secret. |
Base network policy for sandboxes in a project using Oracle-managed outbound networking. |
|
Base outbound networking configuration for sandboxes in a project. |
|
CP-owned configuration and management lifecycle of a Project-owned Sandbox warm pool. |
|
Sandbox Warm Pool Collection. |
|
Read-only default warm-pool capacities for managed-runtime sandboxes. |
|
Best-effort runtime observations returned by GetSandboxWarmPool. |
|
CP-owned configuration and management lifecycle summary. |
|
The auto scaling configuration for the Hosted Application. |
|
The Schedule configuration of a VectorStoreConnector to trigger a File Sync Operation. |
|
The scheduled UNIX cron definition. |
|
The interval schedule config. |
|
Represents a database schema identified by name. |
|
|
Array of database schemas or database objects included in the enrichment pipeline for data sources connected via an OCI Database Tools connection. |
Array of database schemas or other database objects to include in enrichment pipeline. |
|
A Semantic Store is a container resource of semantic records, with controllable enrichment refresh and synchronization policy. |
|
Results of a SemanticStore list. |
|
Model selection configuration for SemanticStore enrichment. |
|
Summary information for a SemanticStore. |
|
The data to set the state of an API key item. |
|
Configuration settings for short-term memory optimization. |
|
Container/artifact configuration for the deployment. |
|
Standard strategy settings for long-term memory. |
|
The type of service-managed storage. |
|
The TFEW training method hyperparameters. |
|
The text generation model metrics of the fine-tuning process. |
|
The fine-tuning method and hyperparameters used for fine-tuning a custom model. |
|
The data to update an API key. |
|
The data to update a dedicated AI cluster. |
|
The data to update an endpoint. |
|
|
The details required to update a Generative AI private endpoint. |
The data to update a GenerativeAiProject. |
|
Common update properties shared by HostedApplication and HostedApplicationIam. |
|
The data to update a hosted application. |
|
The data to update a hosted application. |
|
The data to update a hosted deployment. |
|
The data to update an imported model. |
|
The data to update a custom model. |
|
The data to update a routing profile. |
|
Update Sandbox Warm Pool Details. |
|
The data to update a SemanticStore. |
|
The data to update a VectorStoreConnector. |
|
The Vanilla training method hyperparameters. |
|
A VectorStore Connector offers a lightweight and configurable mechanism to continuously synchronize data from external systems into the VectorStore at scale. |
|
Results of a VectorStoreConnector search. |
|
The VectorStoreConnectorFileSync is an operation that carries out the data sync operation between the datasource and the VectorStore. |
|
|
Results of a VectorStoreConnectorFileSync search. |
|
Summary information for a VectorStoreConnectorFileSync |
A Log object that gives the ingestion status of a File from a datasource read by a VectorStoreConnector |
|
|
Results of a VectorStoreConnector Ingestion Log search. |
File synchronization statistics for a VectorStoreConnector. |
|
Summary information for a VectorStoreConnector |
|
An asynchronous work request. |
|
An error encountered while performing an operation that is tracked by this work request. |
|
A list of work request errors. |
|
The log message from performing an operation that is tracked by this work request. |
|
A list of work request logs. |
|
The resource created or operated on by a work request. |
|
Summary information about an asynchronous work request. |
|
A list of work requests. |