Profile(mapping=None, *, ignore_unknown_fields=False, **kwargs)A profile containing information about a model deployment.
Attributes |
|
|---|---|
| Name | Description |
model_server_info |
google.cloud.gkerecommender_v1.types.ModelServerInfo
Output only. The model server configuration. Use GkeInferenceQuickstart.FetchProfiles to find valid configurations. |
accelerator_type |
str
Output only. The accelerator type. Expected format: nvidia-h100-80gb.
|
tpu_topology |
str
Output only. The TPU topology (if applicable). |
instance_type |
str
Output only. The instance type. Expected format: a2-highgpu-1g.
|
resources_used |
google.cloud.gkerecommender_v1.types.ResourcesUsed
Output only. The resources used by the model deployment. |
performance_stats |
MutableSequence[google.cloud.gkerecommender_v1.types.PerformanceStats]
Output only. The performance statistics for this profile. |