Classes
GkeInferenceQuickstartClient
GKE Inference Quickstart (GIQ) service provides profiles with performance metrics for popular models and model servers across multiple accelerators.
GkeInferenceQuickstartConnection
The GkeInferenceQuickstartConnection object for GkeInferenceQuickstartClient.
GkeInferenceQuickstartConnectionIdempotencyPolicy
GkeInferenceQuickstartLimitedErrorCountRetryPolicy
A retry policy for GkeInferenceQuickstartConnection based on counting errors.
GkeInferenceQuickstartLimitedTimeRetryPolicy
A retry policy for GkeInferenceQuickstartConnection based on elapsed time.
GkeInferenceQuickstartRetryPolicy
The retry policy for GkeInferenceQuickstartConnection.
Structs
GkeInferenceQuickstartBackoffPolicyOption
Use with google::cloud::Options to configure the backoff policy.
GkeInferenceQuickstartConnectionIdempotencyPolicyOption
Use with google::cloud::Options to configure which operations are retried.
GkeInferenceQuickstartRetryPolicyOption
Use with google::cloud::Options to configure the retry policy.
Functions
MakeGkeInferenceQuickstartConnection(Options)
A factory function to construct an object of type GkeInferenceQuickstartConnection.
The returned connection object should not be used directly; instead it should be passed as an argument to the constructor of GkeInferenceQuickstartClient.
The optional options argument may be used to configure aspects of the returned GkeInferenceQuickstartConnection. Expected options are any of the types in the following option lists:
| Parameter | |
|---|---|
| Name | Description |
options |
Options(optional) Configure the |
| Returns | |
|---|---|
| Type | Description |
std::shared_ptr< GkeInferenceQuickstartConnection > |
|
MakeDefaultGkeInferenceQuickstartConnectionIdempotencyPolicy()
| Returns | |
|---|---|
| Type | Description |
std::unique_ptr< GkeInferenceQuickstartConnectionIdempotencyPolicy > |
|