pulumi-gcp 8.40.0a1754721948__py3-none-any.whl → 8.40.0a1754951145__py3-none-any.whl

pulumi_gcp/vertex/ai_endpoint_with_model_garden_deployment.py

@@ -0,0 +1,940 @@
+# coding=utf-8
+# *** WARNING: this file was generated by pulumi-language-python. ***
+# *** Do not edit by hand unless you're certain you know what you are doing! ***
+
+import builtins as _builtins
+import warnings
+import sys
+import pulumi
+import pulumi.runtime
+from typing import Any, Mapping, Optional, Sequence, Union, overload
+if sys.version_info >= (3, 11):
+    from typing import NotRequired, TypedDict, TypeAlias
+else:
+    from typing_extensions import NotRequired, TypedDict, TypeAlias
+from .. import _utilities
+from . import outputs
+from ._inputs import *
+
+__all__ = ['AiEndpointWithModelGardenDeploymentArgs', 'AiEndpointWithModelGardenDeployment']
+
+@pulumi.input_type
+class AiEndpointWithModelGardenDeploymentArgs:
+    def __init__(__self__, *,
+                 location: pulumi.Input[_builtins.str],
+                 deploy_config: Optional[pulumi.Input['AiEndpointWithModelGardenDeploymentDeployConfigArgs']] = None,
+                 endpoint_config: Optional[pulumi.Input['AiEndpointWithModelGardenDeploymentEndpointConfigArgs']] = None,
+                 hugging_face_model_id: Optional[pulumi.Input[_builtins.str]] = None,
+                 model_config: Optional[pulumi.Input['AiEndpointWithModelGardenDeploymentModelConfigArgs']] = None,
+                 project: Optional[pulumi.Input[_builtins.str]] = None,
+                 publisher_model_name: Optional[pulumi.Input[_builtins.str]] = None):
+        """
+        The set of arguments for constructing a AiEndpointWithModelGardenDeployment resource.
+        :param pulumi.Input[_builtins.str] location: Resource ID segment making up resource `location`. It identifies the resource within its parent collection as described in https://google.aip.dev/122.
+        :param pulumi.Input['AiEndpointWithModelGardenDeploymentDeployConfigArgs'] deploy_config: The deploy config to use for the deployment.
+               Structure is documented below.
+        :param pulumi.Input['AiEndpointWithModelGardenDeploymentEndpointConfigArgs'] endpoint_config: The endpoint config to use for the deployment.
+               Structure is documented below.
+        :param pulumi.Input[_builtins.str] hugging_face_model_id: The Hugging Face model to deploy.
+               Format: Hugging Face model ID like `google/gemma-2-2b-it`.
+        :param pulumi.Input['AiEndpointWithModelGardenDeploymentModelConfigArgs'] model_config: The model config to use for the deployment.
+               Structure is documented below.
+        :param pulumi.Input[_builtins.str] project: The ID of the project in which the resource belongs.
+               If it is not provided, the provider project is used.
+        :param pulumi.Input[_builtins.str] publisher_model_name: The Model Garden model to deploy.
+               Format:
+               `publishers/{publisher}/models/{publisher_model}@{version_id}`, or
+               `publishers/hf-{hugging-face-author}/models/{hugging-face-model-name}@001`.
+        """
+        pulumi.set(__self__, "location", location)
+        if deploy_config is not None:
+            pulumi.set(__self__, "deploy_config", deploy_config)
+        if endpoint_config is not None:
+            pulumi.set(__self__, "endpoint_config", endpoint_config)
+        if hugging_face_model_id is not None:
+            pulumi.set(__self__, "hugging_face_model_id", hugging_face_model_id)
+        if model_config is not None:
+            pulumi.set(__self__, "model_config", model_config)
+        if project is not None:
+            pulumi.set(__self__, "project", project)
+        if publisher_model_name is not None:
+            pulumi.set(__self__, "publisher_model_name", publisher_model_name)
+
+    @_builtins.property
+    @pulumi.getter
+    def location(self) -> pulumi.Input[_builtins.str]:
+        """
+        Resource ID segment making up resource `location`. It identifies the resource within its parent collection as described in https://google.aip.dev/122.
+        """
+        return pulumi.get(self, "location")
+
+    @location.setter
+    def location(self, value: pulumi.Input[_builtins.str]):
+        pulumi.set(self, "location", value)
+
+    @_builtins.property
+    @pulumi.getter(name="deployConfig")
+    def deploy_config(self) -> Optional[pulumi.Input['AiEndpointWithModelGardenDeploymentDeployConfigArgs']]:
+        """
+        The deploy config to use for the deployment.
+        Structure is documented below.
+        """
+        return pulumi.get(self, "deploy_config")
+
+    @deploy_config.setter
+    def deploy_config(self, value: Optional[pulumi.Input['AiEndpointWithModelGardenDeploymentDeployConfigArgs']]):
+        pulumi.set(self, "deploy_config", value)
+
+    @_builtins.property
+    @pulumi.getter(name="endpointConfig")
+    def endpoint_config(self) -> Optional[pulumi.Input['AiEndpointWithModelGardenDeploymentEndpointConfigArgs']]:
+        """
+        The endpoint config to use for the deployment.
+        Structure is documented below.
+        """
+        return pulumi.get(self, "endpoint_config")
+
+    @endpoint_config.setter
+    def endpoint_config(self, value: Optional[pulumi.Input['AiEndpointWithModelGardenDeploymentEndpointConfigArgs']]):
+        pulumi.set(self, "endpoint_config", value)
+
+    @_builtins.property
+    @pulumi.getter(name="huggingFaceModelId")
+    def hugging_face_model_id(self) -> Optional[pulumi.Input[_builtins.str]]:
+        """
+        The Hugging Face model to deploy.
+        Format: Hugging Face model ID like `google/gemma-2-2b-it`.
+        """
+        return pulumi.get(self, "hugging_face_model_id")
+
+    @hugging_face_model_id.setter
+    def hugging_face_model_id(self, value: Optional[pulumi.Input[_builtins.str]]):
+        pulumi.set(self, "hugging_face_model_id", value)
+
+    @_builtins.property
+    @pulumi.getter(name="modelConfig")
+    def model_config(self) -> Optional[pulumi.Input['AiEndpointWithModelGardenDeploymentModelConfigArgs']]:
+        """
+        The model config to use for the deployment.
+        Structure is documented below.
+        """
+        return pulumi.get(self, "model_config")
+
+    @model_config.setter
+    def model_config(self, value: Optional[pulumi.Input['AiEndpointWithModelGardenDeploymentModelConfigArgs']]):
+        pulumi.set(self, "model_config", value)
+
+    @_builtins.property
+    @pulumi.getter
+    def project(self) -> Optional[pulumi.Input[_builtins.str]]:
+        """
+        The ID of the project in which the resource belongs.
+        If it is not provided, the provider project is used.
+        """
+        return pulumi.get(self, "project")
+
+    @project.setter
+    def project(self, value: Optional[pulumi.Input[_builtins.str]]):
+        pulumi.set(self, "project", value)
+
+    @_builtins.property
+    @pulumi.getter(name="publisherModelName")
+    def publisher_model_name(self) -> Optional[pulumi.Input[_builtins.str]]:
+        """
+        The Model Garden model to deploy.
+        Format:
+        `publishers/{publisher}/models/{publisher_model}@{version_id}`, or
+        `publishers/hf-{hugging-face-author}/models/{hugging-face-model-name}@001`.
+        """
+        return pulumi.get(self, "publisher_model_name")
+
+    @publisher_model_name.setter
+    def publisher_model_name(self, value: Optional[pulumi.Input[_builtins.str]]):
+        pulumi.set(self, "publisher_model_name", value)
+
+
+@pulumi.input_type
+class _AiEndpointWithModelGardenDeploymentState:
+    def __init__(__self__, *,
+                 deploy_config: Optional[pulumi.Input['AiEndpointWithModelGardenDeploymentDeployConfigArgs']] = None,
+                 deployed_model_display_name: Optional[pulumi.Input[_builtins.str]] = None,
+                 deployed_model_id: Optional[pulumi.Input[_builtins.str]] = None,
+                 endpoint: Optional[pulumi.Input[_builtins.str]] = None,
+                 endpoint_config: Optional[pulumi.Input['AiEndpointWithModelGardenDeploymentEndpointConfigArgs']] = None,
+                 hugging_face_model_id: Optional[pulumi.Input[_builtins.str]] = None,
+                 location: Optional[pulumi.Input[_builtins.str]] = None,
+                 model_config: Optional[pulumi.Input['AiEndpointWithModelGardenDeploymentModelConfigArgs']] = None,
+                 project: Optional[pulumi.Input[_builtins.str]] = None,
+                 publisher_model_name: Optional[pulumi.Input[_builtins.str]] = None):
+        """
+        Input properties used for looking up and filtering AiEndpointWithModelGardenDeployment resources.
+        :param pulumi.Input['AiEndpointWithModelGardenDeploymentDeployConfigArgs'] deploy_config: The deploy config to use for the deployment.
+               Structure is documented below.
+        :param pulumi.Input[_builtins.str] deployed_model_display_name: Output only. The display name assigned to the model deployed to the endpoint.
+               This is not required to delete the resource but is used for debug logging.
+        :param pulumi.Input[_builtins.str] deployed_model_id: Output only. The unique numeric ID that Vertex AI assigns to the model at the time it is deployed to the endpoint.
+               It is required to undeploy the model from the endpoint during resource deletion as described in
+               https://cloud.google.com/vertex-ai/docs/reference/rest/v1/projects.locations.endpoints/undeployModel.
+        :param pulumi.Input[_builtins.str] endpoint: Resource ID segment making up resource `endpoint`. It identifies the resource within its parent collection as described in https://google.aip.dev/122.
+        :param pulumi.Input['AiEndpointWithModelGardenDeploymentEndpointConfigArgs'] endpoint_config: The endpoint config to use for the deployment.
+               Structure is documented below.
+        :param pulumi.Input[_builtins.str] hugging_face_model_id: The Hugging Face model to deploy.
+               Format: Hugging Face model ID like `google/gemma-2-2b-it`.
+        :param pulumi.Input[_builtins.str] location: Resource ID segment making up resource `location`. It identifies the resource within its parent collection as described in https://google.aip.dev/122.
+        :param pulumi.Input['AiEndpointWithModelGardenDeploymentModelConfigArgs'] model_config: The model config to use for the deployment.
+               Structure is documented below.
+        :param pulumi.Input[_builtins.str] project: The ID of the project in which the resource belongs.
+               If it is not provided, the provider project is used.
+        :param pulumi.Input[_builtins.str] publisher_model_name: The Model Garden model to deploy.
+               Format:
+               `publishers/{publisher}/models/{publisher_model}@{version_id}`, or
+               `publishers/hf-{hugging-face-author}/models/{hugging-face-model-name}@001`.
+        """
+        if deploy_config is not None:
+            pulumi.set(__self__, "deploy_config", deploy_config)
+        if deployed_model_display_name is not None:
+            pulumi.set(__self__, "deployed_model_display_name", deployed_model_display_name)
+        if deployed_model_id is not None:
+            pulumi.set(__self__, "deployed_model_id", deployed_model_id)
+        if endpoint is not None:
+            pulumi.set(__self__, "endpoint", endpoint)
+        if endpoint_config is not None:
+            pulumi.set(__self__, "endpoint_config", endpoint_config)
+        if hugging_face_model_id is not None:
+            pulumi.set(__self__, "hugging_face_model_id", hugging_face_model_id)
+        if location is not None:
+            pulumi.set(__self__, "location", location)
+        if model_config is not None:
+            pulumi.set(__self__, "model_config", model_config)
+        if project is not None:
+            pulumi.set(__self__, "project", project)
+        if publisher_model_name is not None:
+            pulumi.set(__self__, "publisher_model_name", publisher_model_name)
+
+    @_builtins.property
+    @pulumi.getter(name="deployConfig")
+    def deploy_config(self) -> Optional[pulumi.Input['AiEndpointWithModelGardenDeploymentDeployConfigArgs']]:
+        """
+        The deploy config to use for the deployment.
+        Structure is documented below.
+        """
+        return pulumi.get(self, "deploy_config")
+
+    @deploy_config.setter
+    def deploy_config(self, value: Optional[pulumi.Input['AiEndpointWithModelGardenDeploymentDeployConfigArgs']]):
+        pulumi.set(self, "deploy_config", value)
+
+    @_builtins.property
+    @pulumi.getter(name="deployedModelDisplayName")
+    def deployed_model_display_name(self) -> Optional[pulumi.Input[_builtins.str]]:
+        """
+        Output only. The display name assigned to the model deployed to the endpoint.
+        This is not required to delete the resource but is used for debug logging.
+        """
+        return pulumi.get(self, "deployed_model_display_name")
+
+    @deployed_model_display_name.setter
+    def deployed_model_display_name(self, value: Optional[pulumi.Input[_builtins.str]]):
+        pulumi.set(self, "deployed_model_display_name", value)
+
+    @_builtins.property
+    @pulumi.getter(name="deployedModelId")
+    def deployed_model_id(self) -> Optional[pulumi.Input[_builtins.str]]:
+        """
+        Output only. The unique numeric ID that Vertex AI assigns to the model at the time it is deployed to the endpoint.
+        It is required to undeploy the model from the endpoint during resource deletion as described in
+        https://cloud.google.com/vertex-ai/docs/reference/rest/v1/projects.locations.endpoints/undeployModel.
+        """
+        return pulumi.get(self, "deployed_model_id")
+
+    @deployed_model_id.setter
+    def deployed_model_id(self, value: Optional[pulumi.Input[_builtins.str]]):
+        pulumi.set(self, "deployed_model_id", value)
+
+    @_builtins.property
+    @pulumi.getter
+    def endpoint(self) -> Optional[pulumi.Input[_builtins.str]]:
+        """
+        Resource ID segment making up resource `endpoint`. It identifies the resource within its parent collection as described in https://google.aip.dev/122.
+        """
+        return pulumi.get(self, "endpoint")
+
+    @endpoint.setter
+    def endpoint(self, value: Optional[pulumi.Input[_builtins.str]]):
+        pulumi.set(self, "endpoint", value)
+
+    @_builtins.property
+    @pulumi.getter(name="endpointConfig")
+    def endpoint_config(self) -> Optional[pulumi.Input['AiEndpointWithModelGardenDeploymentEndpointConfigArgs']]:
+        """
+        The endpoint config to use for the deployment.
+        Structure is documented below.
+        """
+        return pulumi.get(self, "endpoint_config")
+
+    @endpoint_config.setter
+    def endpoint_config(self, value: Optional[pulumi.Input['AiEndpointWithModelGardenDeploymentEndpointConfigArgs']]):
+        pulumi.set(self, "endpoint_config", value)
+
+    @_builtins.property
+    @pulumi.getter(name="huggingFaceModelId")
+    def hugging_face_model_id(self) -> Optional[pulumi.Input[_builtins.str]]:
+        """
+        The Hugging Face model to deploy.
+        Format: Hugging Face model ID like `google/gemma-2-2b-it`.
+        """
+        return pulumi.get(self, "hugging_face_model_id")
+
+    @hugging_face_model_id.setter
+    def hugging_face_model_id(self, value: Optional[pulumi.Input[_builtins.str]]):
+        pulumi.set(self, "hugging_face_model_id", value)
+
+    @_builtins.property
+    @pulumi.getter
+    def location(self) -> Optional[pulumi.Input[_builtins.str]]:
+        """
+        Resource ID segment making up resource `location`. It identifies the resource within its parent collection as described in https://google.aip.dev/122.
+        """
+        return pulumi.get(self, "location")
+
+    @location.setter
+    def location(self, value: Optional[pulumi.Input[_builtins.str]]):
+        pulumi.set(self, "location", value)
+
+    @_builtins.property
+    @pulumi.getter(name="modelConfig")
+    def model_config(self) -> Optional[pulumi.Input['AiEndpointWithModelGardenDeploymentModelConfigArgs']]:
+        """
+        The model config to use for the deployment.
+        Structure is documented below.
+        """
+        return pulumi.get(self, "model_config")
+
+    @model_config.setter
+    def model_config(self, value: Optional[pulumi.Input['AiEndpointWithModelGardenDeploymentModelConfigArgs']]):
+        pulumi.set(self, "model_config", value)
+
+    @_builtins.property
+    @pulumi.getter
+    def project(self) -> Optional[pulumi.Input[_builtins.str]]:
+        """
+        The ID of the project in which the resource belongs.
+        If it is not provided, the provider project is used.
+        """
+        return pulumi.get(self, "project")
+
+    @project.setter
+    def project(self, value: Optional[pulumi.Input[_builtins.str]]):
+        pulumi.set(self, "project", value)
+
+    @_builtins.property
+    @pulumi.getter(name="publisherModelName")
+    def publisher_model_name(self) -> Optional[pulumi.Input[_builtins.str]]:
+        """
+        The Model Garden model to deploy.
+        Format:
+        `publishers/{publisher}/models/{publisher_model}@{version_id}`, or
+        `publishers/hf-{hugging-face-author}/models/{hugging-face-model-name}@001`.
+        """
+        return pulumi.get(self, "publisher_model_name")
+
+    @publisher_model_name.setter
+    def publisher_model_name(self, value: Optional[pulumi.Input[_builtins.str]]):
+        pulumi.set(self, "publisher_model_name", value)
+
+
+@pulumi.type_token("gcp:vertex/aiEndpointWithModelGardenDeployment:AiEndpointWithModelGardenDeployment")
+class AiEndpointWithModelGardenDeployment(pulumi.CustomResource):
+    @overload
+    def __init__(__self__,
+                 resource_name: str,
+                 opts: Optional[pulumi.ResourceOptions] = None,
+                 deploy_config: Optional[pulumi.Input[Union['AiEndpointWithModelGardenDeploymentDeployConfigArgs', 'AiEndpointWithModelGardenDeploymentDeployConfigArgsDict']]] = None,
+                 endpoint_config: Optional[pulumi.Input[Union['AiEndpointWithModelGardenDeploymentEndpointConfigArgs', 'AiEndpointWithModelGardenDeploymentEndpointConfigArgsDict']]] = None,
+                 hugging_face_model_id: Optional[pulumi.Input[_builtins.str]] = None,
+                 location: Optional[pulumi.Input[_builtins.str]] = None,
+                 model_config: Optional[pulumi.Input[Union['AiEndpointWithModelGardenDeploymentModelConfigArgs', 'AiEndpointWithModelGardenDeploymentModelConfigArgsDict']]] = None,
+                 project: Optional[pulumi.Input[_builtins.str]] = None,
+                 publisher_model_name: Optional[pulumi.Input[_builtins.str]] = None,
+                 __props__=None):
+        """
+        Create an Endpoint and deploy a Model Garden model to it.
+
+        To get more information about EndpointWithModelGardenDeployment, see:
+
+        * [API documentation](https://cloud.google.com/vertex-ai/docs/reference/rest/v1/projects.locations/deploy)
+        * How-to Guides
+            * [Overview of Model Garden](https://cloud.google.com/vertex-ai/generative-ai/docs/model-garden/explore-models)
+            * [Overview of self-deployed models](https://cloud.google.com/vertex-ai/generative-ai/docs/model-garden/self-deployed-models)
+            * [Use models in Model Garden](https://cloud.google.com/vertex-ai/generative-ai/docs/model-garden/use-models)
+
+        ## Example Usage
+
+        ### Vertex Ai Deploy Basic
+
+        ```python
+        import pulumi
+        import pulumi_gcp as gcp
+
+        deploy = gcp.vertex.AiEndpointWithModelGardenDeployment("deploy",
+            publisher_model_name="publishers/google/models/paligemma@paligemma-224-float32",
+            location="us-central1",
+            model_config={
+                "accept_eula": True,
+            })
+        ```
+        ### Vertex Ai Deploy Huggingface Model
+
+        ```python
+        import pulumi
+        import pulumi_gcp as gcp
+
+        deploy = gcp.vertex.AiEndpointWithModelGardenDeployment("deploy",
+            hugging_face_model_id="Qwen/Qwen3-0.6B",
+            location="us-central1",
+            model_config={
+                "accept_eula": True,
+            })
+        ```
+        ### Vertex Ai Deploy With Configs
+
+        ```python
+        import pulumi
+        import pulumi_gcp as gcp
+
+        deploy = gcp.vertex.AiEndpointWithModelGardenDeployment("deploy",
+            publisher_model_name="publishers/google/models/paligemma@paligemma-224-float32",
+            location="us-central1",
+            model_config={
+                "accept_eula": True,
+            },
+            deploy_config={
+                "dedicated_resources": {
+                    "machine_spec": {
+                        "machine_type": "g2-standard-16",
+                        "accelerator_type": "NVIDIA_L4",
+                        "accelerator_count": 1,
+                    },
+                    "min_replica_count": 1,
+                },
+            })
+        ```
+        ### Vertex Ai Deploy Multiple Models In Parallel
+
+        ```python
+        import pulumi
+        import pulumi_gcp as gcp
+
+        deploy_gemma_11_2b_it = gcp.vertex.AiEndpointWithModelGardenDeployment("deploy-gemma-1_1-2b-it",
+            publisher_model_name="publishers/google/models/gemma@gemma-1.1-2b-it",
+            location="us-central1",
+            model_config={
+                "accept_eula": True,
+            },
+            deploy_config={
+                "dedicated_resources": {
+                    "machine_spec": {
+                        "machine_type": "g2-standard-12",
+                        "accelerator_type": "us-central1",
+                        "accelerator_count": 1,
+                    },
+                    "min_replica_count": 1,
+                },
+            })
+        deploy_qwen3_06b = gcp.vertex.AiEndpointWithModelGardenDeployment("deploy-qwen3-0_6b",
+            hugging_face_model_id="Qwen/Qwen3-0.6B",
+            location="us-central1",
+            model_config={
+                "accept_eula": True,
+            },
+            deploy_config={
+                "dedicated_resources": {
+                    "machine_spec": {
+                        "machine_type": "g2-standard-12",
+                        "accelerator_type": "NVIDIA_L4",
+                        "accelerator_count": 1,
+                    },
+                    "min_replica_count": 1,
+                },
+            })
+        deploy_llama_32_1b = gcp.vertex.AiEndpointWithModelGardenDeployment("deploy-llama-3_2-1b",
+            publisher_model_name="publishers/meta/models/llama3-2@llama-3.2-1b",
+            location="us-central1",
+            model_config={
+                "accept_eula": True,
+            },
+            deploy_config={
+                "dedicated_resources": {
+                    "machine_spec": {
+                        "machine_type": "g2-standard-12",
+                        "accelerator_type": "NVIDIA_L4",
+                        "accelerator_count": 1,
+                    },
+                    "min_replica_count": 1,
+                },
+            })
+        ```
+        ### Vertex Ai Deploy Multiple Models In Sequence
+
+        ```python
+        import pulumi
+        import pulumi_gcp as gcp
+
+        deploy_gemma_11_2b_it = gcp.vertex.AiEndpointWithModelGardenDeployment("deploy-gemma-1_1-2b-it",
+            publisher_model_name="publishers/google/models/gemma@gemma-1.1-2b-it",
+            location="us-central1",
+            model_config={
+                "accept_eula": True,
+            },
+            deploy_config={
+                "dedicated_resources": {
+                    "machine_spec": {
+                        "machine_type": "g2-standard-12",
+                        "accelerator_type": "NVIDIA_L4",
+                        "accelerator_count": 1,
+                    },
+                    "min_replica_count": 1,
+                },
+            })
+        deploy_qwen3_06b = gcp.vertex.AiEndpointWithModelGardenDeployment("deploy-qwen3-0_6b",
+            hugging_face_model_id="Qwen/Qwen3-0.6B",
+            location="us-central1",
+            model_config={
+                "accept_eula": True,
+            },
+            deploy_config={
+                "dedicated_resources": {
+                    "machine_spec": {
+                        "machine_type": "g2-standard-12",
+                        "accelerator_type": "NVIDIA_L4",
+                        "accelerator_count": 1,
+                    },
+                    "min_replica_count": 1,
+                },
+            },
+            opts = pulumi.ResourceOptions(depends_on=[deploy_gemma_11_2b_it]))
+        deploy_llama_32_1b = gcp.vertex.AiEndpointWithModelGardenDeployment("deploy-llama-3_2-1b",
+            publisher_model_name="publishers/meta/models/llama3-2@llama-3.2-1b",
+            location="us-central1",
+            model_config={
+                "accept_eula": True,
+            },
+            deploy_config={
+                "dedicated_resources": {
+                    "machine_spec": {
+                        "machine_type": "g2-standard-12",
+                        "accelerator_type": "NVIDIA_L4",
+                        "accelerator_count": 1,
+                    },
+                    "min_replica_count": 1,
+                },
+            },
+            opts = pulumi.ResourceOptions(depends_on=[deploy_qwen3_06b]))
+        ```
+
+        ## Import
+
+        This resource does not support import.
+
+        :param str resource_name: The name of the resource.
+        :param pulumi.ResourceOptions opts: Options for the resource.
+        :param pulumi.Input[Union['AiEndpointWithModelGardenDeploymentDeployConfigArgs', 'AiEndpointWithModelGardenDeploymentDeployConfigArgsDict']] deploy_config: The deploy config to use for the deployment.
+               Structure is documented below.
+        :param pulumi.Input[Union['AiEndpointWithModelGardenDeploymentEndpointConfigArgs', 'AiEndpointWithModelGardenDeploymentEndpointConfigArgsDict']] endpoint_config: The endpoint config to use for the deployment.
+               Structure is documented below.
+        :param pulumi.Input[_builtins.str] hugging_face_model_id: The Hugging Face model to deploy.
+               Format: Hugging Face model ID like `google/gemma-2-2b-it`.
+        :param pulumi.Input[_builtins.str] location: Resource ID segment making up resource `location`. It identifies the resource within its parent collection as described in https://google.aip.dev/122.
+        :param pulumi.Input[Union['AiEndpointWithModelGardenDeploymentModelConfigArgs', 'AiEndpointWithModelGardenDeploymentModelConfigArgsDict']] model_config: The model config to use for the deployment.
+               Structure is documented below.
+        :param pulumi.Input[_builtins.str] project: The ID of the project in which the resource belongs.
+               If it is not provided, the provider project is used.
+        :param pulumi.Input[_builtins.str] publisher_model_name: The Model Garden model to deploy.
+               Format:
+               `publishers/{publisher}/models/{publisher_model}@{version_id}`, or
+               `publishers/hf-{hugging-face-author}/models/{hugging-face-model-name}@001`.
+        """
+        ...
+    @overload
+    def __init__(__self__,
+                 resource_name: str,
+                 args: AiEndpointWithModelGardenDeploymentArgs,
+                 opts: Optional[pulumi.ResourceOptions] = None):
+        """
+        Create an Endpoint and deploy a Model Garden model to it.
+
+        To get more information about EndpointWithModelGardenDeployment, see:
+
+        * [API documentation](https://cloud.google.com/vertex-ai/docs/reference/rest/v1/projects.locations/deploy)
+        * How-to Guides
+            * [Overview of Model Garden](https://cloud.google.com/vertex-ai/generative-ai/docs/model-garden/explore-models)
+            * [Overview of self-deployed models](https://cloud.google.com/vertex-ai/generative-ai/docs/model-garden/self-deployed-models)
+            * [Use models in Model Garden](https://cloud.google.com/vertex-ai/generative-ai/docs/model-garden/use-models)
+
+        ## Example Usage
+
+        ### Vertex Ai Deploy Basic
+
+        ```python
+        import pulumi
+        import pulumi_gcp as gcp
+
+        deploy = gcp.vertex.AiEndpointWithModelGardenDeployment("deploy",
+            publisher_model_name="publishers/google/models/paligemma@paligemma-224-float32",
+            location="us-central1",
+            model_config={
+                "accept_eula": True,
+            })
+        ```
+        ### Vertex Ai Deploy Huggingface Model
+
+        ```python
+        import pulumi
+        import pulumi_gcp as gcp
+
+        deploy = gcp.vertex.AiEndpointWithModelGardenDeployment("deploy",
+            hugging_face_model_id="Qwen/Qwen3-0.6B",
+            location="us-central1",
+            model_config={
+                "accept_eula": True,
+            })
+        ```
+        ### Vertex Ai Deploy With Configs
+
+        ```python
+        import pulumi
+        import pulumi_gcp as gcp
+
+        deploy = gcp.vertex.AiEndpointWithModelGardenDeployment("deploy",
+            publisher_model_name="publishers/google/models/paligemma@paligemma-224-float32",
+            location="us-central1",
+            model_config={
+                "accept_eula": True,
+            },
+            deploy_config={
+                "dedicated_resources": {
+                    "machine_spec": {
+                        "machine_type": "g2-standard-16",
+                        "accelerator_type": "NVIDIA_L4",
+                        "accelerator_count": 1,
+                    },
+                    "min_replica_count": 1,
+                },
+            })
+        ```
+        ### Vertex Ai Deploy Multiple Models In Parallel
+
+        ```python
+        import pulumi
+        import pulumi_gcp as gcp
+
+        deploy_gemma_11_2b_it = gcp.vertex.AiEndpointWithModelGardenDeployment("deploy-gemma-1_1-2b-it",
+            publisher_model_name="publishers/google/models/gemma@gemma-1.1-2b-it",
+            location="us-central1",
+            model_config={
+                "accept_eula": True,
+            },
+            deploy_config={
+                "dedicated_resources": {
+                    "machine_spec": {
+                        "machine_type": "g2-standard-12",
+                        "accelerator_type": "us-central1",
+                        "accelerator_count": 1,
+                    },
+                    "min_replica_count": 1,
+                },
+            })
+        deploy_qwen3_06b = gcp.vertex.AiEndpointWithModelGardenDeployment("deploy-qwen3-0_6b",
+            hugging_face_model_id="Qwen/Qwen3-0.6B",
+            location="us-central1",
+            model_config={
+                "accept_eula": True,
+            },
+            deploy_config={
+                "dedicated_resources": {
+                    "machine_spec": {
+                        "machine_type": "g2-standard-12",
+                        "accelerator_type": "NVIDIA_L4",
+                        "accelerator_count": 1,
+                    },
+                    "min_replica_count": 1,
+                },
+            })
+        deploy_llama_32_1b = gcp.vertex.AiEndpointWithModelGardenDeployment("deploy-llama-3_2-1b",
+            publisher_model_name="publishers/meta/models/llama3-2@llama-3.2-1b",
+            location="us-central1",
+            model_config={
+                "accept_eula": True,
+            },
+            deploy_config={
+                "dedicated_resources": {
+                    "machine_spec": {
+                        "machine_type": "g2-standard-12",
+                        "accelerator_type": "NVIDIA_L4",
+                        "accelerator_count": 1,
+                    },
+                    "min_replica_count": 1,
+                },
+            })
+        ```
+        ### Vertex Ai Deploy Multiple Models In Sequence
+
+        ```python
+        import pulumi
+        import pulumi_gcp as gcp
+
+        deploy_gemma_11_2b_it = gcp.vertex.AiEndpointWithModelGardenDeployment("deploy-gemma-1_1-2b-it",
+            publisher_model_name="publishers/google/models/gemma@gemma-1.1-2b-it",
+            location="us-central1",
+            model_config={
+                "accept_eula": True,
+            },
+            deploy_config={
+                "dedicated_resources": {
+                    "machine_spec": {
+                        "machine_type": "g2-standard-12",
+                        "accelerator_type": "NVIDIA_L4",
+                        "accelerator_count": 1,
+                    },
+                    "min_replica_count": 1,
+                },
+            })
+        deploy_qwen3_06b = gcp.vertex.AiEndpointWithModelGardenDeployment("deploy-qwen3-0_6b",
+            hugging_face_model_id="Qwen/Qwen3-0.6B",
+            location="us-central1",
+            model_config={
+                "accept_eula": True,
+            },
+            deploy_config={
+                "dedicated_resources": {
+                    "machine_spec": {
+                        "machine_type": "g2-standard-12",
+                        "accelerator_type": "NVIDIA_L4",
+                        "accelerator_count": 1,
+                    },
+                    "min_replica_count": 1,
+                },
+            },
+            opts = pulumi.ResourceOptions(depends_on=[deploy_gemma_11_2b_it]))
+        deploy_llama_32_1b = gcp.vertex.AiEndpointWithModelGardenDeployment("deploy-llama-3_2-1b",
+            publisher_model_name="publishers/meta/models/llama3-2@llama-3.2-1b",
+            location="us-central1",
+            model_config={
+                "accept_eula": True,
+            },
+            deploy_config={
+                "dedicated_resources": {
+                    "machine_spec": {
+                        "machine_type": "g2-standard-12",
+                        "accelerator_type": "NVIDIA_L4",
+                        "accelerator_count": 1,
+                    },
+                    "min_replica_count": 1,
+                },
+            },
+            opts = pulumi.ResourceOptions(depends_on=[deploy_qwen3_06b]))
+        ```
+
+        ## Import
+
+        This resource does not support import.
+
+        :param str resource_name: The name of the resource.
+        :param AiEndpointWithModelGardenDeploymentArgs args: The arguments to use to populate this resource's properties.
+        :param pulumi.ResourceOptions opts: Options for the resource.
+        """
+        ...
+    def __init__(__self__, resource_name: str, *args, **kwargs):
+        resource_args, opts = _utilities.get_resource_args_opts(AiEndpointWithModelGardenDeploymentArgs, pulumi.ResourceOptions, *args, **kwargs)
+        if resource_args is not None:
+            __self__._internal_init(resource_name, opts, **resource_args.__dict__)
+        else:
+            __self__._internal_init(resource_name, *args, **kwargs)
+
+    def _internal_init(__self__,
+                 resource_name: str,
+                 opts: Optional[pulumi.ResourceOptions] = None,
+                 deploy_config: Optional[pulumi.Input[Union['AiEndpointWithModelGardenDeploymentDeployConfigArgs', 'AiEndpointWithModelGardenDeploymentDeployConfigArgsDict']]] = None,
+                 endpoint_config: Optional[pulumi.Input[Union['AiEndpointWithModelGardenDeploymentEndpointConfigArgs', 'AiEndpointWithModelGardenDeploymentEndpointConfigArgsDict']]] = None,
+                 hugging_face_model_id: Optional[pulumi.Input[_builtins.str]] = None,
+                 location: Optional[pulumi.Input[_builtins.str]] = None,
+                 model_config: Optional[pulumi.Input[Union['AiEndpointWithModelGardenDeploymentModelConfigArgs', 'AiEndpointWithModelGardenDeploymentModelConfigArgsDict']]] = None,
+                 project: Optional[pulumi.Input[_builtins.str]] = None,
+                 publisher_model_name: Optional[pulumi.Input[_builtins.str]] = None,
+                 __props__=None):
+        opts = pulumi.ResourceOptions.merge(_utilities.get_resource_opts_defaults(), opts)
+        if not isinstance(opts, pulumi.ResourceOptions):
+            raise TypeError('Expected resource options to be a ResourceOptions instance')
+        if opts.id is None:
+            if __props__ is not None:
+                raise TypeError('__props__ is only valid when passed in combination with a valid opts.id to get an existing resource')
+            __props__ = AiEndpointWithModelGardenDeploymentArgs.__new__(AiEndpointWithModelGardenDeploymentArgs)
+
+            __props__.__dict__["deploy_config"] = deploy_config
+            __props__.__dict__["endpoint_config"] = endpoint_config
+            __props__.__dict__["hugging_face_model_id"] = hugging_face_model_id
+            if location is None and not opts.urn:
+                raise TypeError("Missing required property 'location'")
+            __props__.__dict__["location"] = location
+            __props__.__dict__["model_config"] = model_config
+            __props__.__dict__["project"] = project
+            __props__.__dict__["publisher_model_name"] = publisher_model_name
+            __props__.__dict__["deployed_model_display_name"] = None
+            __props__.__dict__["deployed_model_id"] = None
+            __props__.__dict__["endpoint"] = None
+        super(AiEndpointWithModelGardenDeployment, __self__).__init__(
+            'gcp:vertex/aiEndpointWithModelGardenDeployment:AiEndpointWithModelGardenDeployment',
+            resource_name,
+            __props__,
+            opts)
+
+    @staticmethod
+    def get(resource_name: str,
+            id: pulumi.Input[str],
+            opts: Optional[pulumi.ResourceOptions] = None,
+            deploy_config: Optional[pulumi.Input[Union['AiEndpointWithModelGardenDeploymentDeployConfigArgs', 'AiEndpointWithModelGardenDeploymentDeployConfigArgsDict']]] = None,
+            deployed_model_display_name: Optional[pulumi.Input[_builtins.str]] = None,
+            deployed_model_id: Optional[pulumi.Input[_builtins.str]] = None,
+            endpoint: Optional[pulumi.Input[_builtins.str]] = None,
+            endpoint_config: Optional[pulumi.Input[Union['AiEndpointWithModelGardenDeploymentEndpointConfigArgs', 'AiEndpointWithModelGardenDeploymentEndpointConfigArgsDict']]] = None,
+            hugging_face_model_id: Optional[pulumi.Input[_builtins.str]] = None,
+            location: Optional[pulumi.Input[_builtins.str]] = None,
+            model_config: Optional[pulumi.Input[Union['AiEndpointWithModelGardenDeploymentModelConfigArgs', 'AiEndpointWithModelGardenDeploymentModelConfigArgsDict']]] = None,
+            project: Optional[pulumi.Input[_builtins.str]] = None,
+            publisher_model_name: Optional[pulumi.Input[_builtins.str]] = None) -> 'AiEndpointWithModelGardenDeployment':
+        """
+        Get an existing AiEndpointWithModelGardenDeployment resource's state with the given name, id, and optional extra
+        properties used to qualify the lookup.
+
+        :param str resource_name: The unique name of the resulting resource.
+        :param pulumi.Input[str] id: The unique provider ID of the resource to lookup.
+        :param pulumi.ResourceOptions opts: Options for the resource.
+        :param pulumi.Input[Union['AiEndpointWithModelGardenDeploymentDeployConfigArgs', 'AiEndpointWithModelGardenDeploymentDeployConfigArgsDict']] deploy_config: The deploy config to use for the deployment.
+               Structure is documented below.
+        :param pulumi.Input[_builtins.str] deployed_model_display_name: Output only. The display name assigned to the model deployed to the endpoint.
+               This is not required to delete the resource but is used for debug logging.
+        :param pulumi.Input[_builtins.str] deployed_model_id: Output only. The unique numeric ID that Vertex AI assigns to the model at the time it is deployed to the endpoint.
+               It is required to undeploy the model from the endpoint during resource deletion as described in
+               https://cloud.google.com/vertex-ai/docs/reference/rest/v1/projects.locations.endpoints/undeployModel.
+        :param pulumi.Input[_builtins.str] endpoint: Resource ID segment making up resource `endpoint`. It identifies the resource within its parent collection as described in https://google.aip.dev/122.
+        :param pulumi.Input[Union['AiEndpointWithModelGardenDeploymentEndpointConfigArgs', 'AiEndpointWithModelGardenDeploymentEndpointConfigArgsDict']] endpoint_config: The endpoint config to use for the deployment.
+               Structure is documented below.
+        :param pulumi.Input[_builtins.str] hugging_face_model_id: The Hugging Face model to deploy.
+               Format: Hugging Face model ID like `google/gemma-2-2b-it`.
+        :param pulumi.Input[_builtins.str] location: Resource ID segment making up resource `location`. It identifies the resource within its parent collection as described in https://google.aip.dev/122.
+        :param pulumi.Input[Union['AiEndpointWithModelGardenDeploymentModelConfigArgs', 'AiEndpointWithModelGardenDeploymentModelConfigArgsDict']] model_config: The model config to use for the deployment.
+               Structure is documented below.
+        :param pulumi.Input[_builtins.str] project: The ID of the project in which the resource belongs.
+               If it is not provided, the provider project is used.
+        :param pulumi.Input[_builtins.str] publisher_model_name: The Model Garden model to deploy.
+               Format:
+               `publishers/{publisher}/models/{publisher_model}@{version_id}`, or
+               `publishers/hf-{hugging-face-author}/models/{hugging-face-model-name}@001`.
+        """
+        opts = pulumi.ResourceOptions.merge(opts, pulumi.ResourceOptions(id=id))
+
+        __props__ = _AiEndpointWithModelGardenDeploymentState.__new__(_AiEndpointWithModelGardenDeploymentState)
+
+        __props__.__dict__["deploy_config"] = deploy_config
+        __props__.__dict__["deployed_model_display_name"] = deployed_model_display_name
+        __props__.__dict__["deployed_model_id"] = deployed_model_id
+        __props__.__dict__["endpoint"] = endpoint
+        __props__.__dict__["endpoint_config"] = endpoint_config
+        __props__.__dict__["hugging_face_model_id"] = hugging_face_model_id
+        __props__.__dict__["location"] = location
+        __props__.__dict__["model_config"] = model_config
+        __props__.__dict__["project"] = project
+        __props__.__dict__["publisher_model_name"] = publisher_model_name
+        return AiEndpointWithModelGardenDeployment(resource_name, opts=opts, __props__=__props__)
+
+    @_builtins.property
+    @pulumi.getter(name="deployConfig")
+    def deploy_config(self) -> pulumi.Output[Optional['outputs.AiEndpointWithModelGardenDeploymentDeployConfig']]:
+        """
+        The deploy config to use for the deployment.
+        Structure is documented below.
+        """
+        return pulumi.get(self, "deploy_config")
+
+    @_builtins.property
+    @pulumi.getter(name="deployedModelDisplayName")
+    def deployed_model_display_name(self) -> pulumi.Output[_builtins.str]:
+        """
+        Output only. The display name assigned to the model deployed to the endpoint.
+        This is not required to delete the resource but is used for debug logging.
+        """
+        return pulumi.get(self, "deployed_model_display_name")
+
+    @_builtins.property
+    @pulumi.getter(name="deployedModelId")
+    def deployed_model_id(self) -> pulumi.Output[_builtins.str]:
+        """
+        Output only. The unique numeric ID that Vertex AI assigns to the model at the time it is deployed to the endpoint.
+        It is required to undeploy the model from the endpoint during resource deletion as described in
+        https://cloud.google.com/vertex-ai/docs/reference/rest/v1/projects.locations.endpoints/undeployModel.
+        """
+        return pulumi.get(self, "deployed_model_id")
+
+    @_builtins.property
+    @pulumi.getter
+    def endpoint(self) -> pulumi.Output[_builtins.str]:
+        """
+        Resource ID segment making up resource `endpoint`. It identifies the resource within its parent collection as described in https://google.aip.dev/122.
+        """
+        return pulumi.get(self, "endpoint")
+
+    @_builtins.property
+    @pulumi.getter(name="endpointConfig")
+    def endpoint_config(self) -> pulumi.Output[Optional['outputs.AiEndpointWithModelGardenDeploymentEndpointConfig']]:
+        """
+        The endpoint config to use for the deployment.
+        Structure is documented below.
+        """
+        return pulumi.get(self, "endpoint_config")
+
+    @_builtins.property
+    @pulumi.getter(name="huggingFaceModelId")
+    def hugging_face_model_id(self) -> pulumi.Output[Optional[_builtins.str]]:
+        """
+        The Hugging Face model to deploy.
+        Format: Hugging Face model ID like `google/gemma-2-2b-it`.
+        """
+        return pulumi.get(self, "hugging_face_model_id")
+
+    @_builtins.property
+    @pulumi.getter
+    def location(self) -> pulumi.Output[_builtins.str]:
+        """
+        Resource ID segment making up resource `location`. It identifies the resource within its parent collection as described in https://google.aip.dev/122.
+        """
+        return pulumi.get(self, "location")
+
+    @_builtins.property
+    @pulumi.getter(name="modelConfig")
+    def model_config(self) -> pulumi.Output[Optional['outputs.AiEndpointWithModelGardenDeploymentModelConfig']]:
+        """
+        The model config to use for the deployment.
+        Structure is documented below.
+        """
+        return pulumi.get(self, "model_config")
+
+    @_builtins.property
+    @pulumi.getter
+    def project(self) -> pulumi.Output[_builtins.str]:
+        """
+        The ID of the project in which the resource belongs.
+        If it is not provided, the provider project is used.
+        """
+        return pulumi.get(self, "project")
+
+    @_builtins.property
+    @pulumi.getter(name="publisherModelName")
+    def publisher_model_name(self) -> pulumi.Output[Optional[_builtins.str]]:
+        """
+        The Model Garden model to deploy.
+        Format:
+        `publishers/{publisher}/models/{publisher_model}@{version_id}`, or
+        `publishers/hf-{hugging-face-author}/models/{hugging-face-model-name}@001`.
+        """
+        return pulumi.get(self, "publisher_model_name")
+