google-apis-aiplatform_v1beta1 0.35.0 → 0.37.0

data/lib/google/apis/aiplatform_v1beta1/classes.rb

@@ -474,7 +474,7 @@ module Google
         end
       end
       
-      # Create API error message for Vertex Pipeline. Next Id: 3.
+      # Create API error message for Vertex Pipeline.
       class CloudAiPlatformCommonCreatePipelineJobApiErrorDetail
         include Google::Apis::Core::Hashable
       
@@ -840,7 +840,8 @@ module Google
         end
       end
       
-      # The generic reusable api auth config.
+      # The generic reusable api auth config. Deprecated. Please use AuthConfig (
+      # google/cloud/aiplatform/master/auth.proto) instead.
       class GoogleCloudAiplatformV1beta1ApiAuth
         include Google::Apis::Core::Hashable
       
@@ -1310,15 +1311,22 @@ module Google
       class GoogleCloudAiplatformV1beta1AuthConfigApiKeyConfig
         include Google::Apis::Core::Hashable
       
-        # Required. The name of the SecretManager secret version resource storing the
+        # Optional. The name of the SecretManager secret version resource storing the
         # API key. Format: `projects/`project`/secrets/`secrete`/versions/`version`` -
-        # If specified, the `secretmanager.versions.access` permission should be granted
-        # to Vertex AI Extension Service Agent (https://cloud.google.com/vertex-ai/docs/
-        # general/access-control#service-agents) on the specified resource.
+        # If both `api_key_secret` and `api_key_string` are specified, this field takes
+        # precedence over `api_key_string`. - If specified, the `secretmanager.versions.
+        # access` permission should be granted to Vertex AI Extension Service Agent (
+        # https://cloud.google.com/vertex-ai/docs/general/access-control#service-agents)
+        # on the specified resource.
         # Corresponds to the JSON property `apiKeySecret`
         # @return [String]
         attr_accessor :api_key_secret
       
+        # Optional. The API key to be used in the request directly.
+        # Corresponds to the JSON property `apiKeyString`
+        # @return [String]
+        attr_accessor :api_key_string
+      
         # Required. The location of the API key.
         # Corresponds to the JSON property `httpElementLocation`
         # @return [String]
@@ -1337,6 +1345,7 @@ module Google
         # Update properties of this object
         def update!(**args)
           @api_key_secret = args[:api_key_secret] if args.key?(:api_key_secret)
+          @api_key_string = args[:api_key_string] if args.key?(:api_key_string)
           @http_element_location = args[:http_element_location] if args.key?(:http_element_location)
           @name = args[:name] if args.key?(:name)
         end
@@ -1460,22 +1469,22 @@ module Google
       class GoogleCloudAiplatformV1beta1AutomaticResources
         include Google::Apis::Core::Hashable
       
-        # Immutable. The maximum number of replicas this DeployedModel may be deployed
-        # on when the traffic against it increases. If the requested value is too large,
-        # the deployment will error, but if deployment succeeds then the ability to
-        # scale the model to that many replicas is guaranteed (barring service outages).
-        # If traffic against the DeployedModel increases beyond what its replicas at
-        # maximum may handle, a portion of the traffic will be dropped. If this value is
-        # not provided, a no upper bound for scaling under heavy traffic will be assume,
-        # though Vertex AI may be unable to scale beyond certain replica number.
+        # Immutable. The maximum number of replicas that may be deployed on when the
+        # traffic against it increases. If the requested value is too large, the
+        # deployment will error, but if deployment succeeds then the ability to scale to
+        # that many replicas is guaranteed (barring service outages). If traffic
+        # increases beyond what its replicas at maximum may handle, a portion of the
+        # traffic will be dropped. If this value is not provided, a no upper bound for
+        # scaling under heavy traffic will be assume, though Vertex AI may be unable to
+        # scale beyond certain replica number.
         # Corresponds to the JSON property `maxReplicaCount`
         # @return [Fixnum]
         attr_accessor :max_replica_count
       
-        # Immutable. The minimum number of replicas this DeployedModel will be always
-        # deployed on. If traffic against it increases, it may dynamically be deployed
-        # onto more replicas up to max_replica_count, and as traffic decreases, some of
-        # these extra replicas may be freed. If the requested value is too large, the
+        # Immutable. The minimum number of replicas that will be always deployed on. If
+        # traffic against it increases, it may dynamically be deployed onto more
+        # replicas up to max_replica_count, and as traffic decreases, some of these
+        # extra replicas may be freed. If the requested value is too large, the
         # deployment will error.
         # Corresponds to the JSON property `minReplicaCount`
         # @return [Fixnum]
@@ -1492,6 +1501,48 @@ module Google
         end
       end
       
+      # The configs for autorater. This is applicable to both EvaluateInstances and
+      # EvaluateDataset.
+      class GoogleCloudAiplatformV1beta1AutoraterConfig
+        include Google::Apis::Core::Hashable
+      
+        # Optional. The fully qualified name of the publisher model or tuned autorater
+        # endpoint to use. Publisher model format: `projects/`project`/locations/`
+        # location`/publishers/*/models/*` Tuned model endpoint format: `projects/`
+        # project`/locations/`location`/endpoints/`endpoint``
+        # Corresponds to the JSON property `autoraterModel`
+        # @return [String]
+        attr_accessor :autorater_model
+      
+        # Optional. Whether to flip the candidate and baseline responses. This is only
+        # applicable to the pairwise metric. If enabled, also provide PairwiseMetricSpec.
+        # candidate_response_field_name and PairwiseMetricSpec.
+        # baseline_response_field_name. When rendering PairwiseMetricSpec.
+        # metric_prompt_template, the candidate and baseline fields will be flipped for
+        # half of the samples to reduce bias.
+        # Corresponds to the JSON property `flipEnabled`
+        # @return [Boolean]
+        attr_accessor :flip_enabled
+        alias_method :flip_enabled?, :flip_enabled
+      
+        # Optional. Number of samples for each instance in the dataset. If not specified,
+        # the default is 4. Minimum value is 1, maximum value is 32.
+        # Corresponds to the JSON property `samplingCount`
+        # @return [Fixnum]
+        attr_accessor :sampling_count
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @autorater_model = args[:autorater_model] if args.key?(:autorater_model)
+          @flip_enabled = args[:flip_enabled] if args.key?(:flip_enabled)
+          @sampling_count = args[:sampling_count] if args.key?(:sampling_count)
+        end
+      end
+      
       # The metric specification that defines the target resource utilization (CPU
       # utilization, accelerator's duty cycle, and so on) for calculating the desired
       # replica count.
@@ -2738,7 +2789,7 @@ module Google
         end
       end
       
-      # Content blob. It's preferred to send as text directly rather than raw bytes.
+      # Content blob.
       class GoogleCloudAiplatformV1beta1Blob
         include Google::Apis::Core::Hashable
       
@@ -2748,6 +2799,13 @@ module Google
         # @return [String]
         attr_accessor :data
       
+        # Optional. Display name of the blob. Used to provide a label or filename to
+        # distinguish blobs. This field is only returned in PromptMessage for prompt
+        # management. It is not currently used in the Gemini GenerateContent calls.
+        # Corresponds to the JSON property `displayName`
+        # @return [String]
+        attr_accessor :display_name
+      
         # Required. The IANA standard MIME type of the source data.
         # Corresponds to the JSON property `mimeType`
         # @return [String]
@@ -2760,6 +2818,7 @@ module Google
         # Update properties of this object
         def update!(**args)
           @data = args[:data] if args.key?(:data)
+          @display_name = args[:display_name] if args.key?(:display_name)
           @mime_type = args[:mime_type] if args.key?(:mime_type)
         end
       end
@@ -2844,7 +2903,7 @@ module Google
         # @return [Array<Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1Content>]
         attr_accessor :contents
       
-        # Output only. Creatation time of the cache entry.
+        # Output only. Creation time of the cache entry.
         # Corresponds to the JSON property `createTime`
         # @return [String]
         attr_accessor :create_time
@@ -2861,8 +2920,9 @@ module Google
         # @return [String]
         attr_accessor :expire_time
       
-        # Immutable. The name of the publisher model to use for cached content. Format:
-        # projects/`project`/locations/`location`/publishers/`publisher`/models/`model`
+        # Immutable. The name of the `Model` to use for cached content. Currently, only
+        # the published Gemini base models are supported, in form of projects/`PROJECT`/
+        # locations/`LOCATION`/publishers/google/models/`MODEL`
         # Corresponds to the JSON property `model`
         # @return [String]
         attr_accessor :model
@@ -4152,6 +4212,11 @@ module Google
       class GoogleCloudAiplatformV1beta1CountTokensResponse
         include Google::Apis::Core::Hashable
       
+        # Output only. List of modalities that were processed in the request input.
+        # Corresponds to the JSON property `promptTokensDetails`
+        # @return [Array<Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1ModalityTokenCount>]
+        attr_accessor :prompt_tokens_details
+      
         # The total number of billable characters counted across all instances from the
         # request.
         # Corresponds to the JSON property `totalBillableCharacters`
@@ -4169,6 +4234,7 @@ module Google
       
         # Update properties of this object
         def update!(**args)
+          @prompt_tokens_details = args[:prompt_tokens_details] if args.key?(:prompt_tokens_details)
           @total_billable_characters = args[:total_billable_characters] if args.key?(:total_billable_characters)
           @total_tokens = args[:total_tokens] if args.key?(:total_tokens)
         end
@@ -5504,7 +5570,8 @@ module Google
         # @return [String]
         attr_accessor :model_reference
       
-        # Output only. Identifier. The resource name of the Dataset.
+        # Output only. Identifier. The resource name of the Dataset. Format: `projects/`
+        # project`/locations/`location`/datasets/`dataset``
         # Corresponds to the JSON property `name`
         # @return [String]
         attr_accessor :name
@@ -5752,7 +5819,9 @@ module Google
         # @return [String]
         attr_accessor :model_reference
       
-        # Output only. Identifier. The resource name of the DatasetVersion.
+        # Output only. Identifier. The resource name of the DatasetVersion. Format: `
+        # projects/`project`/locations/`location`/datasets/`dataset`/datasetVersions/`
+        # dataset_version``
         # Corresponds to the JSON property `name`
         # @return [String]
         attr_accessor :name
@@ -5793,8 +5862,8 @@ module Google
         end
       end
       
-      # A description of resources that are dedicated to a DeployedModel, and that
-      # need a higher degree of manual configuration.
+      # A description of resources that are dedicated to a DeployedModel or
+      # DeployedIndex, and that need a higher degree of manual configuration.
       class GoogleCloudAiplatformV1beta1DedicatedResources
         include Google::Apis::Core::Hashable
       
@@ -5820,36 +5889,33 @@ module Google
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1MachineSpec]
         attr_accessor :machine_spec
       
-        # Immutable. The maximum number of replicas this DeployedModel may be deployed
-        # on when the traffic against it increases. If the requested value is too large,
-        # the deployment will error, but if deployment succeeds then the ability to
-        # scale the model to that many replicas is guaranteed (barring service outages).
-        # If traffic against the DeployedModel increases beyond what its replicas at
-        # maximum may handle, a portion of the traffic will be dropped. If this value is
-        # not provided, will use min_replica_count as the default value. The value of
-        # this field impacts the charge against Vertex CPU and GPU quotas. Specifically,
-        # you will be charged for (max_replica_count * number of cores in the selected
-        # machine type) and (max_replica_count * number of GPUs per replica in the
-        # selected machine type).
+        # Immutable. The maximum number of replicas that may be deployed on when the
+        # traffic against it increases. If the requested value is too large, the
+        # deployment will error, but if deployment succeeds then the ability to scale to
+        # that many replicas is guaranteed (barring service outages). If traffic
+        # increases beyond what its replicas at maximum may handle, a portion of the
+        # traffic will be dropped. If this value is not provided, will use
+        # min_replica_count as the default value. The value of this field impacts the
+        # charge against Vertex CPU and GPU quotas. Specifically, you will be charged
+        # for (max_replica_count * number of cores in the selected machine type) and (
+        # max_replica_count * number of GPUs per replica in the selected machine type).
         # Corresponds to the JSON property `maxReplicaCount`
         # @return [Fixnum]
         attr_accessor :max_replica_count
       
-        # Required. Immutable. The minimum number of machine replicas this DeployedModel
-        # will be always deployed on. This value must be greater than or equal to 1. If
-        # traffic against the DeployedModel increases, it may dynamically be deployed
-        # onto more replicas, and as traffic decreases, some of these extra replicas may
-        # be freed.
+        # Required. Immutable. The minimum number of machine replicas that will be
+        # always deployed on. This value must be greater than or equal to 1. If traffic
+        # increases, it may dynamically be deployed onto more replicas, and as traffic
+        # decreases, some of these extra replicas may be freed.
         # Corresponds to the JSON property `minReplicaCount`
         # @return [Fixnum]
         attr_accessor :min_replica_count
       
         # Optional. Number of required available replicas for the deployment to succeed.
-        # This field is only needed when partial model deployment/mutation is desired.
-        # If set, the model deploy/mutate operation will succeed once
-        # available_replica_count reaches required_replica_count, and the rest of the
-        # replicas will be retried. If not set, the default required_replica_count will
-        # be min_replica_count.
+        # This field is only needed when partial deployment/mutation is desired. If set,
+        # the deploy/mutate operation will succeed once available_replica_count reaches
+        # required_replica_count, and the rest of the replicas will be retried. If not
+        # set, the default required_replica_count will be min_replica_count.
         # Corresponds to the JSON property `requiredReplicaCount`
         # @return [Fixnum]
         attr_accessor :required_replica_count
@@ -6259,8 +6325,8 @@ module Google
         attr_accessor :accept_eula
         alias_method :accept_eula?, :accept_eula
       
-        # A description of resources that are dedicated to a DeployedModel, and that
-        # need a higher degree of manual configuration.
+        # A description of resources that are dedicated to a DeployedModel or
+        # DeployedIndex, and that need a higher degree of manual configuration.
         # Corresponds to the JSON property `dedicatedResources`
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1DedicatedResources]
         attr_accessor :dedicated_resources
@@ -6277,9 +6343,11 @@ module Google
         # @return [String]
         attr_accessor :hugging_face_access_token
       
-        # Required. The name of the PublisherModel resource. Format: `publishers/`
-        # publisher`/models/`publisher_model`@`version_id``, or `publishers/hf-`hugging-
-        # face-author`/models/`hugging-face-model-name`@001`
+        # Required. The model to deploy. Format: 1. `publishers/`publisher`/models/`
+        # publisher_model`@`version_id``, or `publishers/hf-`hugging-face-author`/models/
+        # `hugging-face-model-name`@001`. 2. Hugging Face model ID like `google/gemma-2-
+        # 2b-it`. 3. Custom model Google Cloud Storage URI like `gs://bucket`. 4. Custom
+        # model zip file like `https://abc.com/a.zip`.
         # Corresponds to the JSON property `model`
         # @return [String]
         attr_accessor :model
@@ -6340,8 +6408,8 @@ module Google
         # @return [String]
         attr_accessor :create_time
       
-        # A description of resources that are dedicated to a DeployedModel, and that
-        # need a higher degree of manual configuration.
+        # A description of resources that are dedicated to a DeployedModel or
+        # DeployedIndex, and that need a higher degree of manual configuration.
         # Corresponds to the JSON property `dedicatedResources`
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1DedicatedResources]
         attr_accessor :dedicated_resources
@@ -6553,8 +6621,8 @@ module Google
         # @return [String]
         attr_accessor :create_time
       
-        # A description of resources that are dedicated to a DeployedModel, and that
-        # need a higher degree of manual configuration.
+        # A description of resources that are dedicated to a DeployedModel or
+        # DeployedIndex, and that need a higher degree of manual configuration.
         # Corresponds to the JSON property `dedicatedResources`
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1DedicatedResources]
         attr_accessor :dedicated_resources
@@ -6630,6 +6698,11 @@ module Google
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1PrivateEndpoints]
         attr_accessor :private_endpoints
       
+        # Configuration for rolling deployments.
+        # Corresponds to the JSON property `rolloutOptions`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1RolloutOptions]
+        attr_accessor :rollout_options
+      
         # The service account that the DeployedModel's container runs as. Specify the
         # email address of the service account. If this service account is not specified,
         # the container runs as a service account that doesn't have access to the
@@ -6646,6 +6719,11 @@ module Google
         # @return [String]
         attr_accessor :shared_resources
       
+        # Configuration for Speculative Decoding.
+        # Corresponds to the JSON property `speculativeDecodingSpec`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1SpeculativeDecodingSpec]
+        attr_accessor :speculative_decoding_spec
+      
         # Runtime status of the deployed model.
         # Corresponds to the JSON property `status`
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1DeployedModelStatus]
@@ -6676,8 +6754,10 @@ module Google
           @model = args[:model] if args.key?(:model)
           @model_version_id = args[:model_version_id] if args.key?(:model_version_id)
           @private_endpoints = args[:private_endpoints] if args.key?(:private_endpoints)
+          @rollout_options = args[:rollout_options] if args.key?(:rollout_options)
           @service_account = args[:service_account] if args.key?(:service_account)
           @shared_resources = args[:shared_resources] if args.key?(:shared_resources)
+          @speculative_decoding_spec = args[:speculative_decoding_spec] if args.key?(:speculative_decoding_spec)
           @status = args[:status] if args.key?(:status)
           @system_labels = args[:system_labels] if args.key?(:system_labels)
         end
@@ -6749,8 +6829,8 @@ module Google
         # @return [String]
         attr_accessor :create_time
       
-        # A description of resources that are dedicated to a DeployedModel, and that
-        # need a higher degree of manual configuration.
+        # A description of resources that are dedicated to a DeployedModel or
+        # DeployedIndex, and that need a higher degree of manual configuration.
         # Corresponds to the JSON property `dedicatedResources`
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1DedicatedResources]
         attr_accessor :dedicated_resources
@@ -7047,9 +7127,9 @@ module Google
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1DistillationHyperParameters]
         attr_accessor :hyper_parameters
       
-        # Required. Deprecated. A path in a Cloud Storage bucket, which will be treated
-        # as the root output directory of the distillation pipeline. It is used by the
-        # system to generate the paths of output artifacts.
+        # Deprecated. A path in a Cloud Storage bucket, which will be treated as the
+        # root output directory of the distillation pipeline. It is used by the system
+        # to generate the paths of output artifacts.
         # Corresponds to the JSON property `pipelineRootDirectory`
         # @return [String]
         attr_accessor :pipeline_root_directory
@@ -7060,7 +7140,7 @@ module Google
         # @return [String]
         attr_accessor :student_model
       
-        # Required. Cloud Storage path to file containing training dataset for tuning.
+        # Deprecated. Cloud Storage path to file containing training dataset for tuning.
         # The dataset must be formatted as a JSONL file.
         # Corresponds to the JSON property `trainingDatasetUri`
         # @return [String]
@@ -7232,6 +7312,11 @@ module Google
         # @return [String]
         attr_accessor :etag
       
+        # Configuration for GenAiAdvancedFeatures.
+        # Corresponds to the JSON property `genAiAdvancedFeaturesConfig`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1GenAiAdvancedFeaturesConfig]
+        attr_accessor :gen_ai_advanced_features_config
+      
         # The labels with user-defined metadata to organize your Endpoints. Label keys
         # and values can be no longer than 64 characters (Unicode codepoints), can only
         # contain lowercase letters, numeric characters, underscores and dashes.
@@ -7318,6 +7403,7 @@ module Google
           @enable_private_service_connect = args[:enable_private_service_connect] if args.key?(:enable_private_service_connect)
           @encryption_spec = args[:encryption_spec] if args.key?(:encryption_spec)
           @etag = args[:etag] if args.key?(:etag)
+          @gen_ai_advanced_features_config = args[:gen_ai_advanced_features_config] if args.key?(:gen_ai_advanced_features_config)
           @labels = args[:labels] if args.key?(:labels)
           @model_deployment_monitoring_job = args[:model_deployment_monitoring_job] if args.key?(:model_deployment_monitoring_job)
           @name = args[:name] if args.key?(:name)
@@ -7545,10 +7631,54 @@ module Google
         end
       end
       
+      # Request message for EvaluationService.EvaluateDataset.
+      class GoogleCloudAiplatformV1beta1EvaluateDatasetRequest
+        include Google::Apis::Core::Hashable
+      
+        # The configs for autorater. This is applicable to both EvaluateInstances and
+        # EvaluateDataset.
+        # Corresponds to the JSON property `autoraterConfig`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1AutoraterConfig]
+        attr_accessor :autorater_config
+      
+        # The dataset used for evaluation.
+        # Corresponds to the JSON property `dataset`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1EvaluationDataset]
+        attr_accessor :dataset
+      
+        # Required. The metrics used for evaluation.
+        # Corresponds to the JSON property `metrics`
+        # @return [Array<Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1Metric>]
+        attr_accessor :metrics
+      
+        # Config for evaluation output.
+        # Corresponds to the JSON property `outputConfig`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1OutputConfig]
+        attr_accessor :output_config
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @autorater_config = args[:autorater_config] if args.key?(:autorater_config)
+          @dataset = args[:dataset] if args.key?(:dataset)
+          @metrics = args[:metrics] if args.key?(:metrics)
+          @output_config = args[:output_config] if args.key?(:output_config)
+        end
+      end
+      
       # Request message for EvaluationService.EvaluateInstances.
       class GoogleCloudAiplatformV1beta1EvaluateInstancesRequest
         include Google::Apis::Core::Hashable
       
+        # The configs for autorater. This is applicable to both EvaluateInstances and
+        # EvaluateDataset.
+        # Corresponds to the JSON property `autoraterConfig`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1AutoraterConfig]
+        attr_accessor :autorater_config
+      
         # Input for bleu metric.
         # Corresponds to the JSON property `bleuInput`
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1BleuInput]
@@ -7710,6 +7840,7 @@ module Google
       
         # Update properties of this object
         def update!(**args)
+          @autorater_config = args[:autorater_config] if args.key?(:autorater_config)
           @bleu_input = args[:bleu_input] if args.key?(:bleu_input)
           @coherence_input = args[:coherence_input] if args.key?(:coherence_input)
           @comet_input = args[:comet_input] if args.key?(:comet_input)
@@ -8049,6 +8180,31 @@ module Google
         end
       end
       
+      # The dataset used for evaluation.
+      class GoogleCloudAiplatformV1beta1EvaluationDataset
+        include Google::Apis::Core::Hashable
+      
+        # The BigQuery location for the input content.
+        # Corresponds to the JSON property `bigquerySource`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1BigQuerySource]
+        attr_accessor :bigquery_source
+      
+        # The Google Cloud Storage location for the input content.
+        # Corresponds to the JSON property `gcsSource`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1GcsSource]
+        attr_accessor :gcs_source
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @bigquery_source = args[:bigquery_source] if args.key?(:bigquery_source)
+          @gcs_source = args[:gcs_source] if args.key?(:gcs_source)
+        end
+      end
+      
       # An edge describing the relationship between an Artifact and an Execution in a
       # lineage graph.
       class GoogleCloudAiplatformV1beta1Event
@@ -9572,6 +9728,51 @@ module Google
         end
       end
       
+      # Runtime operation information for ModelGardenService.ExportPublisherModel.
+      class GoogleCloudAiplatformV1beta1ExportPublisherModelOperationMetadata
+        include Google::Apis::Core::Hashable
+      
+        # Generic Metadata shared by all operations.
+        # Corresponds to the JSON property `genericMetadata`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1GenericOperationMetadata]
+        attr_accessor :generic_metadata
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @generic_metadata = args[:generic_metadata] if args.key?(:generic_metadata)
+        end
+      end
+      
+      # Response message for ModelGardenService.ExportPublisherModel.
+      class GoogleCloudAiplatformV1beta1ExportPublisherModelResponse
+        include Google::Apis::Core::Hashable
+      
+        # The destination uri of the model weights.
+        # Corresponds to the JSON property `destinationUri`
+        # @return [String]
+        attr_accessor :destination_uri
+      
+        # The name of the PublisherModel resource. Format: `publishers/`publisher`/
+        # models/`publisher_model`@`version_id``
+        # Corresponds to the JSON property `publisherModel`
+        # @return [String]
+        attr_accessor :publisher_model
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @destination_uri = args[:destination_uri] if args.key?(:destination_uri)
+          @publisher_model = args[:publisher_model] if args.key?(:publisher_model)
+        end
+      end
+      
       # Request message for TensorboardService.ExportTensorboardTimeSeriesData.
       class GoogleCloudAiplatformV1beta1ExportTensorboardTimeSeriesDataRequest
         include Google::Apis::Core::Hashable
@@ -12183,6 +12384,14 @@ module Google
       class GoogleCloudAiplatformV1beta1FileData
         include Google::Apis::Core::Hashable
       
+        # Optional. Display name of the file data. Used to provide a label or filename
+        # to distinguish file datas. This field is only returned in PromptMessage for
+        # prompt management. It is not currently used in the Gemini GenerateContent
+        # calls.
+        # Corresponds to the JSON property `displayName`
+        # @return [String]
+        attr_accessor :display_name
+      
         # Required. URI.
         # Corresponds to the JSON property `fileUri`
         # @return [String]
@@ -12199,6 +12408,7 @@ module Google
       
         # Update properties of this object
         def update!(**args)
+          @display_name = args[:display_name] if args.key?(:display_name)
           @file_uri = args[:file_uri] if args.key?(:file_uri)
           @mime_type = args[:mime_type] if args.key?(:mime_type)
         end
@@ -12881,6 +13091,47 @@ module Google
         end
       end
       
+      # Configuration for GenAiAdvancedFeatures.
+      class GoogleCloudAiplatformV1beta1GenAiAdvancedFeaturesConfig
+        include Google::Apis::Core::Hashable
+      
+        # Configuration for Retrieval Augmented Generation feature.
+        # Corresponds to the JSON property `ragConfig`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1GenAiAdvancedFeaturesConfigRagConfig]
+        attr_accessor :rag_config
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @rag_config = args[:rag_config] if args.key?(:rag_config)
+        end
+      end
+      
+      # Configuration for Retrieval Augmented Generation feature.
+      class GoogleCloudAiplatformV1beta1GenAiAdvancedFeaturesConfigRagConfig
+        include Google::Apis::Core::Hashable
+      
+        # If true, enable Retrieval Augmented Generation in ChatCompletion request. Once
+        # enabled, the endpoint will be identified as GenAI endpoint and Arthedain
+        # router will be used.
+        # Corresponds to the JSON property `enableRag`
+        # @return [Boolean]
+        attr_accessor :enable_rag
+        alias_method :enable_rag?, :enable_rag
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @enable_rag = args[:enable_rag] if args.key?(:enable_rag)
+        end
+      end
+      
       # Request message for NotebookInternalService.GenerateAccessToken.
       class GoogleCloudAiplatformV1beta1GenerateAccessTokenRequest
         include Google::Apis::Core::Hashable
@@ -13028,6 +13279,11 @@ module Google
         # @return [Array<Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1Candidate>]
         attr_accessor :candidates
       
+        # Output only. Timestamp when the request is made to the server.
+        # Corresponds to the JSON property `createTime`
+        # @return [String]
+        attr_accessor :create_time
+      
         # Output only. The model version used to generate the response.
         # Corresponds to the JSON property `modelVersion`
         # @return [String]
@@ -13038,6 +13294,12 @@ module Google
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1GenerateContentResponsePromptFeedback]
         attr_accessor :prompt_feedback
       
+        # Output only. response_id is used to identify each response. It is the encoding
+        # of the event_id.
+        # Corresponds to the JSON property `responseId`
+        # @return [String]
+        attr_accessor :response_id
+      
         # Usage metadata about response(s).
         # Corresponds to the JSON property `usageMetadata`
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1GenerateContentResponseUsageMetadata]
@@ -13050,8 +13312,10 @@ module Google
         # Update properties of this object
         def update!(**args)
           @candidates = args[:candidates] if args.key?(:candidates)
+          @create_time = args[:create_time] if args.key?(:create_time)
           @model_version = args[:model_version] if args.key?(:model_version)
           @prompt_feedback = args[:prompt_feedback] if args.key?(:prompt_feedback)
+          @response_id = args[:response_id] if args.key?(:response_id)
           @usage_metadata = args[:usage_metadata] if args.key?(:usage_metadata)
         end
       end
@@ -13091,6 +13355,11 @@ module Google
       class GoogleCloudAiplatformV1beta1GenerateContentResponseUsageMetadata
         include Google::Apis::Core::Hashable
       
+        # Output only. List of modalities of the cached content in the request input.
+        # Corresponds to the JSON property `cacheTokensDetails`
+        # @return [Array<Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1ModalityTokenCount>]
+        attr_accessor :cache_tokens_details
+      
         # Output only. Number of tokens in the cached part in the input (the cached
         # content).
         # Corresponds to the JSON property `cachedContentTokenCount`
@@ -13102,6 +13371,11 @@ module Google
         # @return [Fixnum]
         attr_accessor :candidates_token_count
       
+        # Output only. List of modalities that were returned in the response.
+        # Corresponds to the JSON property `candidatesTokensDetails`
+        # @return [Array<Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1ModalityTokenCount>]
+        attr_accessor :candidates_tokens_details
+      
         # Number of tokens in the request. When `cached_content` is set, this is still
         # the total effective prompt size meaning this includes the number of tokens in
         # the cached content.
@@ -13109,6 +13383,11 @@ module Google
         # @return [Fixnum]
         attr_accessor :prompt_token_count
       
+        # Output only. List of modalities that were processed in the request input.
+        # Corresponds to the JSON property `promptTokensDetails`
+        # @return [Array<Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1ModalityTokenCount>]
+        attr_accessor :prompt_tokens_details
+      
         # Total token count for prompt and response candidates.
         # Corresponds to the JSON property `totalTokenCount`
         # @return [Fixnum]
@@ -13120,9 +13399,12 @@ module Google
       
         # Update properties of this object
         def update!(**args)
+          @cache_tokens_details = args[:cache_tokens_details] if args.key?(:cache_tokens_details)
           @cached_content_token_count = args[:cached_content_token_count] if args.key?(:cached_content_token_count)
           @candidates_token_count = args[:candidates_token_count] if args.key?(:candidates_token_count)
+          @candidates_tokens_details = args[:candidates_tokens_details] if args.key?(:candidates_tokens_details)
           @prompt_token_count = args[:prompt_token_count] if args.key?(:prompt_token_count)
+          @prompt_tokens_details = args[:prompt_tokens_details] if args.key?(:prompt_tokens_details)
           @total_token_count = args[:total_token_count] if args.key?(:total_token_count)
         end
       end
@@ -13189,6 +13471,11 @@ module Google
         # @return [Fixnum]
         attr_accessor :max_output_tokens
       
+        # Optional. If specified, the media resolution specified will be used.
+        # Corresponds to the JSON property `mediaResolution`
+        # @return [String]
+        attr_accessor :media_resolution
+      
         # Optional. Positive penalties.
         # Corresponds to the JSON property `presencePenalty`
         # @return [Float]
@@ -13246,11 +13533,6 @@ module Google
         # @return [Float]
         attr_accessor :temperature
       
-        # Optional. If specified, the token resolution specified will be used.
-        # Corresponds to the JSON property `tokenResolution`
-        # @return [String]
-        attr_accessor :token_resolution
-      
         # Optional. If specified, top-k sampling will be used.
         # Corresponds to the JSON property `topK`
         # @return [Float]
@@ -13272,6 +13554,7 @@ module Google
           @frequency_penalty = args[:frequency_penalty] if args.key?(:frequency_penalty)
           @logprobs = args[:logprobs] if args.key?(:logprobs)
           @max_output_tokens = args[:max_output_tokens] if args.key?(:max_output_tokens)
+          @media_resolution = args[:media_resolution] if args.key?(:media_resolution)
           @presence_penalty = args[:presence_penalty] if args.key?(:presence_penalty)
           @response_logprobs = args[:response_logprobs] if args.key?(:response_logprobs)
           @response_mime_type = args[:response_mime_type] if args.key?(:response_mime_type)
@@ -13282,7 +13565,6 @@ module Google
           @speech_config = args[:speech_config] if args.key?(:speech_config)
           @stop_sequences = args[:stop_sequences] if args.key?(:stop_sequences)
           @temperature = args[:temperature] if args.key?(:temperature)
-          @token_resolution = args[:token_resolution] if args.key?(:token_resolution)
           @top_k = args[:top_k] if args.key?(:top_k)
           @top_p = args[:top_p] if args.key?(:top_p)
         end
@@ -16075,6 +16357,32 @@ module Google
         end
       end
       
+      # Response message for ModelService.ListModelVersionCheckpoints
+      class GoogleCloudAiplatformV1beta1ListModelVersionCheckpointsResponse
+        include Google::Apis::Core::Hashable
+      
+        # List of Model Version checkpoints.
+        # Corresponds to the JSON property `checkpoints`
+        # @return [Array<Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1ModelVersionCheckpoint>]
+        attr_accessor :checkpoints
+      
+        # A token to retrieve the next page of results. Pass to
+        # ListModelVersionCheckpointsRequest.page_token to obtain that page.
+        # Corresponds to the JSON property `nextPageToken`
+        # @return [String]
+        attr_accessor :next_page_token
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @checkpoints = args[:checkpoints] if args.key?(:checkpoints)
+          @next_page_token = args[:next_page_token] if args.key?(:next_page_token)
+        end
+      end
+      
       # Response message for ModelService.ListModelVersions
       class GoogleCloudAiplatformV1beta1ListModelVersionsResponse
         include Google::Apis::Core::Hashable
@@ -17147,6 +17455,58 @@ module Google
         end
       end
       
+      # The metric used for dataset level evaluation.
+      class GoogleCloudAiplatformV1beta1Metric
+        include Google::Apis::Core::Hashable
+      
+        # Optional. The aggregation metrics to use.
+        # Corresponds to the JSON property `aggregationMetrics`
+        # @return [Array<String>]
+        attr_accessor :aggregation_metrics
+      
+        # Spec for bleu score metric - calculates the precision of n-grams in the
+        # prediction as compared to reference - returns a score ranging between 0 to 1.
+        # Corresponds to the JSON property `bleuSpec`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1BleuSpec]
+        attr_accessor :bleu_spec
+      
+        # Spec for exact match metric - returns 1 if prediction and reference exactly
+        # matches, otherwise 0.
+        # Corresponds to the JSON property `exactMatchSpec`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1ExactMatchSpec]
+        attr_accessor :exact_match_spec
+      
+        # Spec for pairwise metric.
+        # Corresponds to the JSON property `pairwiseMetricSpec`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1PairwiseMetricSpec]
+        attr_accessor :pairwise_metric_spec
+      
+        # Spec for pointwise metric.
+        # Corresponds to the JSON property `pointwiseMetricSpec`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1PointwiseMetricSpec]
+        attr_accessor :pointwise_metric_spec
+      
+        # Spec for rouge score metric - calculates the recall of n-grams in prediction
+        # as compared to reference - returns a score ranging between 0 and 1.
+        # Corresponds to the JSON property `rougeSpec`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1RougeSpec]
+        attr_accessor :rouge_spec
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @aggregation_metrics = args[:aggregation_metrics] if args.key?(:aggregation_metrics)
+          @bleu_spec = args[:bleu_spec] if args.key?(:bleu_spec)
+          @exact_match_spec = args[:exact_match_spec] if args.key?(:exact_match_spec)
+          @pairwise_metric_spec = args[:pairwise_metric_spec] if args.key?(:pairwise_metric_spec)
+          @pointwise_metric_spec = args[:pointwise_metric_spec] if args.key?(:pointwise_metric_spec)
+          @rouge_spec = args[:rouge_spec] if args.key?(:rouge_spec)
+        end
+      end
+      
       # Input for MetricX metric.
       class GoogleCloudAiplatformV1beta1MetricxInput
         include Google::Apis::Core::Hashable
@@ -17665,6 +18025,31 @@ module Google
         end
       end
       
+      # Represents token counting info for a single modality.
+      class GoogleCloudAiplatformV1beta1ModalityTokenCount
+        include Google::Apis::Core::Hashable
+      
+        # The modality associated with this token count.
+        # Corresponds to the JSON property `modality`
+        # @return [String]
+        attr_accessor :modality
+      
+        # Number of tokens.
+        # Corresponds to the JSON property `tokenCount`
+        # @return [Fixnum]
+        attr_accessor :token_count
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @modality = args[:modality] if args.key?(:modality)
+          @token_count = args[:token_count] if args.key?(:token_count)
+        end
+      end
+      
       # A trained machine learning Model.
       class GoogleCloudAiplatformV1beta1Model
         include Google::Apis::Core::Hashable
@@ -17694,6 +18079,11 @@ module Google
         # @return [String]
         attr_accessor :create_time
       
+        # The default checkpoint id of a model version.
+        # Corresponds to the JSON property `defaultCheckpointId`
+        # @return [String]
+        attr_accessor :default_checkpoint_id
+      
         # Output only. The pointers to DeployedModels created from this Model. Note that
         # Model could have been deployed to Endpoints in different Locations.
         # Corresponds to the JSON property `deployedModels`
@@ -17908,6 +18298,7 @@ module Google
           @base_model_source = args[:base_model_source] if args.key?(:base_model_source)
           @container_spec = args[:container_spec] if args.key?(:container_spec)
           @create_time = args[:create_time] if args.key?(:create_time)
+          @default_checkpoint_id = args[:default_checkpoint_id] if args.key?(:default_checkpoint_id)
           @deployed_models = args[:deployed_models] if args.key?(:deployed_models)
           @description = args[:description] if args.key?(:description)
           @display_name = args[:display_name] if args.key?(:display_name)
@@ -18110,6 +18501,12 @@ module Google
         # @return [String]
         attr_accessor :image_uri
       
+        # Probe describes a health check to be performed against a container to
+        # determine whether it is alive or ready to receive traffic.
+        # Corresponds to the JSON property `livenessProbe`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1Probe]
+        attr_accessor :liveness_probe
+      
         # Immutable. List of ports to expose from the container. Vertex AI sends any
         # prediction requests that it receives to the first port on this list. Vertex AI
         # also sends [liveness and health checks](https://cloud.google.com/vertex-ai/
@@ -18172,6 +18569,7 @@ module Google
           @health_probe = args[:health_probe] if args.key?(:health_probe)
           @health_route = args[:health_route] if args.key?(:health_route)
           @image_uri = args[:image_uri] if args.key?(:image_uri)
+          @liveness_probe = args[:liveness_probe] if args.key?(:liveness_probe)
           @ports = args[:ports] if args.key?(:ports)
           @predict_route = args[:predict_route] if args.key?(:predict_route)
           @shared_memory_size_mb = args[:shared_memory_size_mb] if args.key?(:shared_memory_size_mb)
@@ -18909,7 +19307,7 @@ module Google
         # model in SavedModel format. * `tf-js` A [TensorFlow.js](https://www.tensorflow.
         # org/js) model that can be used in the browser and in Node.js using JavaScript.
         # * `core-ml` Used for iOS mobile devices. * `custom-trained` A Model that was
-        # uploaded or trained by custom code.
+        # uploaded or trained by custom code. * `genie` A tuned Model Garden model.
         # Corresponds to the JSON property `id`
         # @return [String]
         attr_accessor :id
@@ -18935,6 +19333,17 @@ module Google
         # @return [String]
         attr_accessor :public_model_name
       
+        # Optional. Whether to avoid pulling the model from the HF cache.
+        # Corresponds to the JSON property `skipHfModelCache`
+        # @return [Boolean]
+        attr_accessor :skip_hf_model_cache
+        alias_method :skip_hf_model_cache?, :skip_hf_model_cache
+      
+        # Optional. The model garden source model version ID.
+        # Corresponds to the JSON property `versionId`
+        # @return [String]
+        attr_accessor :version_id
+      
         def initialize(**args)
            update!(**args)
         end
@@ -18942,6 +19351,8 @@ module Google
         # Update properties of this object
         def update!(**args)
           @public_model_name = args[:public_model_name] if args.key?(:public_model_name)
+          @skip_hf_model_cache = args[:skip_hf_model_cache] if args.key?(:skip_hf_model_cache)
+          @version_id = args[:version_id] if args.key?(:version_id)
         end
       end
       
@@ -20594,6 +21005,45 @@ module Google
         end
       end
       
+      # Describes the machine learning model version checkpoint.
+      class GoogleCloudAiplatformV1beta1ModelVersionCheckpoint
+        include Google::Apis::Core::Hashable
+      
+        # The ID of the checkpoint.
+        # Corresponds to the JSON property `checkpointId`
+        # @return [String]
+        attr_accessor :checkpoint_id
+      
+        # The epoch of the checkpoint.
+        # Corresponds to the JSON property `epoch`
+        # @return [Fixnum]
+        attr_accessor :epoch
+      
+        # Identifier. The resource name of the ModelVersionCheckpoint. Format: `projects/
+        # `project`/locations/`location`/models/`model`/versions/`version`/checkpoints/`
+        # checkpoint``
+        # Corresponds to the JSON property `name`
+        # @return [String]
+        attr_accessor :name
+      
+        # The step of the checkpoint.
+        # Corresponds to the JSON property `step`
+        # @return [Fixnum]
+        attr_accessor :step
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @checkpoint_id = args[:checkpoint_id] if args.key?(:checkpoint_id)
+          @epoch = args[:epoch] if args.key?(:epoch)
+          @name = args[:name] if args.key?(:name)
+          @step = args[:step] if args.key?(:step)
+        end
+      end
+      
       # Runtime operation information for IndexEndpointService.MutateDeployedIndex.
       class GoogleCloudAiplatformV1beta1MutateDeployedIndexOperationMetadata
         include Google::Apis::Core::Hashable
@@ -21703,8 +22153,8 @@ module Google
         # @return [String]
         attr_accessor :notebook_runtime_template_resource_name
       
-        # Output only. The Schedule resource name if this job is triggered by one.
-        # Format: `projects/`project_id`/locations/`location`/schedules/`schedule_id``
+        # The Schedule resource name if this job is triggered by one. Format: `projects/`
+        # project_id`/locations/`location`/schedules/`schedule_id``
         # Corresponds to the JSON property `scheduleResourceName`
         # @return [String]
         attr_accessor :schedule_resource_name
@@ -22067,6 +22517,11 @@ module Google
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1ShieldedVmConfig]
         attr_accessor :shielded_vm_config
       
+        # Notebook Software Config.
+        # Corresponds to the JSON property `softwareConfig`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1NotebookSoftwareConfig]
+        attr_accessor :software_config
+      
         # Output only. Timestamp when this NotebookRuntime was most recently updated.
         # Corresponds to the JSON property `updateTime`
         # @return [String]
@@ -22107,6 +22562,7 @@ module Google
           @satisfies_pzs = args[:satisfies_pzs] if args.key?(:satisfies_pzs)
           @service_account = args[:service_account] if args.key?(:service_account)
           @shielded_vm_config = args[:shielded_vm_config] if args.key?(:shielded_vm_config)
+          @software_config = args[:software_config] if args.key?(:software_config)
           @update_time = args[:update_time] if args.key?(:update_time)
           @version = args[:version] if args.key?(:version)
         end
@@ -22225,6 +22681,11 @@ module Google
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1ShieldedVmConfig]
         attr_accessor :shielded_vm_config
       
+        # Notebook Software Config.
+        # Corresponds to the JSON property `softwareConfig`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1NotebookSoftwareConfig]
+        attr_accessor :software_config
+      
         # Output only. Timestamp when this NotebookRuntimeTemplate was most recently
         # updated.
         # Corresponds to the JSON property `updateTime`
@@ -22254,6 +22715,7 @@ module Google
           @notebook_runtime_type = args[:notebook_runtime_type] if args.key?(:notebook_runtime_type)
           @service_account = args[:service_account] if args.key?(:service_account)
           @shielded_vm_config = args[:shielded_vm_config] if args.key?(:shielded_vm_config)
+          @software_config = args[:software_config] if args.key?(:software_config)
           @update_time = args[:update_time] if args.key?(:update_time)
         end
       end
@@ -22277,6 +22739,51 @@ module Google
         end
       end
       
+      # Notebook Software Config.
+      class GoogleCloudAiplatformV1beta1NotebookSoftwareConfig
+        include Google::Apis::Core::Hashable
+      
+        # Optional. Environment variables to be passed to the container. Maximum limit
+        # is 100.
+        # Corresponds to the JSON property `env`
+        # @return [Array<Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1EnvVar>]
+        attr_accessor :env
+      
+        # Post startup script config.
+        # Corresponds to the JSON property `postStartupScriptConfig`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1PostStartupScriptConfig]
+        attr_accessor :post_startup_script_config
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @env = args[:env] if args.key?(:env)
+          @post_startup_script_config = args[:post_startup_script_config] if args.key?(:post_startup_script_config)
+        end
+      end
+      
+      # Config for evaluation output.
+      class GoogleCloudAiplatformV1beta1OutputConfig
+        include Google::Apis::Core::Hashable
+      
+        # The Google Cloud Storage location where the output is to be written to.
+        # Corresponds to the JSON property `gcsDestination`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1GcsDestination]
+        attr_accessor :gcs_destination
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @gcs_destination = args[:gcs_destination] if args.key?(:gcs_destination)
+        end
+      end
+      
       # PSC config that is used to automatically create forwarding rule via
       # ServiceConnectionMap.
       class GoogleCloudAiplatformV1beta1PscAutomationConfig
@@ -22383,18 +22890,36 @@ module Google
       class GoogleCloudAiplatformV1beta1PairwiseMetricSpec
         include Google::Apis::Core::Hashable
       
+        # Optional. The field name of the baseline response.
+        # Corresponds to the JSON property `baselineResponseFieldName`
+        # @return [String]
+        attr_accessor :baseline_response_field_name
+      
+        # Optional. The field name of the candidate response.
+        # Corresponds to the JSON property `candidateResponseFieldName`
+        # @return [String]
+        attr_accessor :candidate_response_field_name
+      
         # Required. Metric prompt template for pairwise metric.
         # Corresponds to the JSON property `metricPromptTemplate`
         # @return [String]
         attr_accessor :metric_prompt_template
       
+        # Optional. System instructions for pairwise metric.
+        # Corresponds to the JSON property `systemInstruction`
+        # @return [String]
+        attr_accessor :system_instruction
+      
         def initialize(**args)
            update!(**args)
         end
       
         # Update properties of this object
         def update!(**args)
+          @baseline_response_field_name = args[:baseline_response_field_name] if args.key?(:baseline_response_field_name)
+          @candidate_response_field_name = args[:candidate_response_field_name] if args.key?(:candidate_response_field_name)
           @metric_prompt_template = args[:metric_prompt_template] if args.key?(:metric_prompt_template)
+          @system_instruction = args[:system_instruction] if args.key?(:system_instruction)
         end
       end
       
@@ -22691,7 +23216,7 @@ module Google
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1FunctionResponse]
         attr_accessor :function_response
       
-        # Content blob. It's preferred to send as text directly rather than raw bytes.
+        # Content blob.
         # Corresponds to the JSON property `inlineData`
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1Blob]
         attr_accessor :inline_data
@@ -22701,6 +23226,12 @@ module Google
         # @return [String]
         attr_accessor :text
       
+        # Output only. Indicates if the part is thought from the model.
+        # Corresponds to the JSON property `thought`
+        # @return [Boolean]
+        attr_accessor :thought
+        alias_method :thought?, :thought
+      
         # Metadata describes the input video content.
         # Corresponds to the JSON property `videoMetadata`
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1VideoMetadata]
@@ -22719,6 +23250,7 @@ module Google
           @function_response = args[:function_response] if args.key?(:function_response)
           @inline_data = args[:inline_data] if args.key?(:inline_data)
           @text = args[:text] if args.key?(:text)
+          @thought = args[:thought] if args.key?(:thought)
           @video_metadata = args[:video_metadata] if args.key?(:video_metadata)
         end
       end
@@ -23785,6 +24317,11 @@ module Google
         # @return [String]
         attr_accessor :metric_prompt_template
       
+        # Optional. System instructions for pointwise metric.
+        # Corresponds to the JSON property `systemInstruction`
+        # @return [String]
+        attr_accessor :system_instruction
+      
         def initialize(**args)
            update!(**args)
         end
@@ -23792,6 +24329,7 @@ module Google
         # Update properties of this object
         def update!(**args)
           @metric_prompt_template = args[:metric_prompt_template] if args.key?(:metric_prompt_template)
+          @system_instruction = args[:system_instruction] if args.key?(:system_instruction)
         end
       end
       
@@ -23815,6 +24353,39 @@ module Google
         end
       end
       
+      # Post startup script config.
+      class GoogleCloudAiplatformV1beta1PostStartupScriptConfig
+        include Google::Apis::Core::Hashable
+      
+        # Optional. Post startup script to run after runtime is started.
+        # Corresponds to the JSON property `postStartupScript`
+        # @return [String]
+        attr_accessor :post_startup_script
+      
+        # Optional. Post startup script behavior that defines download and execution
+        # behavior.
+        # Corresponds to the JSON property `postStartupScriptBehavior`
+        # @return [String]
+        attr_accessor :post_startup_script_behavior
+      
+        # Optional. Post startup script url to download. Example: https://bucket/script.
+        # sh
+        # Corresponds to the JSON property `postStartupScriptUrl`
+        # @return [String]
+        attr_accessor :post_startup_script_url
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @post_startup_script = args[:post_startup_script] if args.key?(:post_startup_script)
+          @post_startup_script_behavior = args[:post_startup_script_behavior] if args.key?(:post_startup_script_behavior)
+          @post_startup_script_url = args[:post_startup_script_url] if args.key?(:post_startup_script_url)
+        end
+      end
+      
       # The configuration for the prebuilt speaker to use.
       class GoogleCloudAiplatformV1beta1PrebuiltVoiceConfig
         include Google::Apis::Core::Hashable
@@ -24221,6 +24792,28 @@ module Google
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1ProbeExecAction]
         attr_accessor :exec
       
+        # Number of consecutive failures before the probe is considered failed. Defaults
+        # to 3. Minimum value is 1. Maps to Kubernetes probe argument 'failureThreshold'.
+        # Corresponds to the JSON property `failureThreshold`
+        # @return [Fixnum]
+        attr_accessor :failure_threshold
+      
+        # GrpcAction checks the health of a container using a gRPC service.
+        # Corresponds to the JSON property `grpc`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1ProbeGrpcAction]
+        attr_accessor :grpc
+      
+        # HttpGetAction describes an action based on HTTP Get requests.
+        # Corresponds to the JSON property `httpGet`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1ProbeHttpGetAction]
+        attr_accessor :http_get
+      
+        # Number of seconds to wait before starting the probe. Defaults to 0. Minimum
+        # value is 0. Maps to Kubernetes probe argument 'initialDelaySeconds'.
+        # Corresponds to the JSON property `initialDelaySeconds`
+        # @return [Fixnum]
+        attr_accessor :initial_delay_seconds
+      
         # How often (in seconds) to perform the probe. Default to 10 seconds. Minimum
         # value is 1. Must be less than timeout_seconds. Maps to Kubernetes probe
         # argument 'periodSeconds'.
@@ -24228,6 +24821,19 @@ module Google
         # @return [Fixnum]
         attr_accessor :period_seconds
       
+        # Number of consecutive successes before the probe is considered successful.
+        # Defaults to 1. Minimum value is 1. Maps to Kubernetes probe argument '
+        # successThreshold'.
+        # Corresponds to the JSON property `successThreshold`
+        # @return [Fixnum]
+        attr_accessor :success_threshold
+      
+        # TcpSocketAction probes the health of a container by opening a TCP socket
+        # connection.
+        # Corresponds to the JSON property `tcpSocket`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1ProbeTcpSocketAction]
+        attr_accessor :tcp_socket
+      
         # Number of seconds after which the probe times out. Defaults to 1 second.
         # Minimum value is 1. Must be greater or equal to period_seconds. Maps to
         # Kubernetes probe argument 'timeoutSeconds'.
@@ -24242,7 +24848,13 @@ module Google
         # Update properties of this object
         def update!(**args)
           @exec = args[:exec] if args.key?(:exec)
+          @failure_threshold = args[:failure_threshold] if args.key?(:failure_threshold)
+          @grpc = args[:grpc] if args.key?(:grpc)
+          @http_get = args[:http_get] if args.key?(:http_get)
+          @initial_delay_seconds = args[:initial_delay_seconds] if args.key?(:initial_delay_seconds)
           @period_seconds = args[:period_seconds] if args.key?(:period_seconds)
+          @success_threshold = args[:success_threshold] if args.key?(:success_threshold)
+          @tcp_socket = args[:tcp_socket] if args.key?(:tcp_socket)
           @timeout_seconds = args[:timeout_seconds] if args.key?(:timeout_seconds)
         end
       end
@@ -24271,6 +24883,133 @@ module Google
         end
       end
       
+      # GrpcAction checks the health of a container using a gRPC service.
+      class GoogleCloudAiplatformV1beta1ProbeGrpcAction
+        include Google::Apis::Core::Hashable
+      
+        # Port number of the gRPC service. Number must be in the range 1 to 65535.
+        # Corresponds to the JSON property `port`
+        # @return [Fixnum]
+        attr_accessor :port
+      
+        # Service is the name of the service to place in the gRPC HealthCheckRequest (
+        # see https://github.com/grpc/grpc/blob/master/doc/health-checking.md). If this
+        # is not specified, the default behavior is defined by gRPC.
+        # Corresponds to the JSON property `service`
+        # @return [String]
+        attr_accessor :service
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @port = args[:port] if args.key?(:port)
+          @service = args[:service] if args.key?(:service)
+        end
+      end
+      
+      # HttpGetAction describes an action based on HTTP Get requests.
+      class GoogleCloudAiplatformV1beta1ProbeHttpGetAction
+        include Google::Apis::Core::Hashable
+      
+        # Host name to connect to, defaults to the model serving container's IP. You
+        # probably want to set "Host" in httpHeaders instead.
+        # Corresponds to the JSON property `host`
+        # @return [String]
+        attr_accessor :host
+      
+        # Custom headers to set in the request. HTTP allows repeated headers.
+        # Corresponds to the JSON property `httpHeaders`
+        # @return [Array<Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1ProbeHttpHeader>]
+        attr_accessor :http_headers
+      
+        # Path to access on the HTTP server.
+        # Corresponds to the JSON property `path`
+        # @return [String]
+        attr_accessor :path
+      
+        # Number of the port to access on the container. Number must be in the range 1
+        # to 65535.
+        # Corresponds to the JSON property `port`
+        # @return [Fixnum]
+        attr_accessor :port
+      
+        # Scheme to use for connecting to the host. Defaults to HTTP. Acceptable values
+        # are "HTTP" or "HTTPS".
+        # Corresponds to the JSON property `scheme`
+        # @return [String]
+        attr_accessor :scheme
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @host = args[:host] if args.key?(:host)
+          @http_headers = args[:http_headers] if args.key?(:http_headers)
+          @path = args[:path] if args.key?(:path)
+          @port = args[:port] if args.key?(:port)
+          @scheme = args[:scheme] if args.key?(:scheme)
+        end
+      end
+      
+      # HttpHeader describes a custom header to be used in HTTP probes
+      class GoogleCloudAiplatformV1beta1ProbeHttpHeader
+        include Google::Apis::Core::Hashable
+      
+        # The header field name. This will be canonicalized upon output, so case-variant
+        # names will be understood as the same header.
+        # Corresponds to the JSON property `name`
+        # @return [String]
+        attr_accessor :name
+      
+        # The header field value
+        # Corresponds to the JSON property `value`
+        # @return [String]
+        attr_accessor :value
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @name = args[:name] if args.key?(:name)
+          @value = args[:value] if args.key?(:value)
+        end
+      end
+      
+      # TcpSocketAction probes the health of a container by opening a TCP socket
+      # connection.
+      class GoogleCloudAiplatformV1beta1ProbeTcpSocketAction
+        include Google::Apis::Core::Hashable
+      
+        # Optional: Host name to connect to, defaults to the model serving container's
+        # IP.
+        # Corresponds to the JSON property `host`
+        # @return [String]
+        attr_accessor :host
+      
+        # Number of the port to access on the container. Number must be in the range 1
+        # to 65535.
+        # Corresponds to the JSON property `port`
+        # @return [Fixnum]
+        attr_accessor :port
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @host = args[:host] if args.key?(:host)
+          @port = args[:port] if args.key?(:port)
+        end
+      end
+      
       # PscAutomatedEndpoints defines the output of the forwarding rule automatically
       # created by each PscAutomationConfig.
       class GoogleCloudAiplatformV1beta1PscAutomatedEndpoints
@@ -24536,8 +25275,8 @@ module Google
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1ModelContainerSpec]
         attr_accessor :container_spec
       
-        # A description of resources that are dedicated to a DeployedModel, and that
-        # need a higher degree of manual configuration.
+        # A description of resources that are dedicated to a DeployedModel or
+        # DeployedIndex, and that need a higher degree of manual configuration.
         # Corresponds to the JSON property `dedicatedResources`
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1DedicatedResources]
         attr_accessor :dedicated_resources
@@ -25867,6 +26606,11 @@ module Google
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1RagEmbeddingModelConfig]
         attr_accessor :rag_embedding_model_config
       
+        # Output only. Number of RagFiles in the RagCorpus.
+        # Corresponds to the JSON property `ragFilesCount`
+        # @return [Fixnum]
+        attr_accessor :rag_files_count
+      
         # Config for the Vector DB to use for RAG.
         # Corresponds to the JSON property `ragVectorDbConfig`
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1RagVectorDbConfig]
@@ -25899,6 +26643,7 @@ module Google
           @display_name = args[:display_name] if args.key?(:display_name)
           @name = args[:name] if args.key?(:name)
           @rag_embedding_model_config = args[:rag_embedding_model_config] if args.key?(:rag_embedding_model_config)
+          @rag_files_count = args[:rag_files_count] if args.key?(:rag_files_count)
           @rag_vector_db_config = args[:rag_vector_db_config] if args.key?(:rag_vector_db_config)
           @update_time = args[:update_time] if args.key?(:update_time)
           @vector_db_config = args[:vector_db_config] if args.key?(:vector_db_config)
@@ -26215,6 +26960,11 @@ module Google
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1RagFileParsingConfigLayoutParser]
         attr_accessor :layout_parser
       
+        # Specifies the advanced parsing for RagFiles.
+        # Corresponds to the JSON property `llmParser`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1RagFileParsingConfigLlmParser]
+        attr_accessor :llm_parser
+      
         # Whether to use advanced PDF parsing.
         # Corresponds to the JSON property `useAdvancedPdfParsing`
         # @return [Boolean]
@@ -26229,6 +26979,7 @@ module Google
         def update!(**args)
           @advanced_parser = args[:advanced_parser] if args.key?(:advanced_parser)
           @layout_parser = args[:layout_parser] if args.key?(:layout_parser)
+          @llm_parser = args[:llm_parser] if args.key?(:llm_parser)
           @use_advanced_pdf_parsing = args[:use_advanced_pdf_parsing] if args.key?(:use_advanced_pdf_parsing)
         end
       end
@@ -26286,6 +27037,41 @@ module Google
         end
       end
       
+      # Specifies the advanced parsing for RagFiles.
+      class GoogleCloudAiplatformV1beta1RagFileParsingConfigLlmParser
+        include Google::Apis::Core::Hashable
+      
+        # The prompt to use for parsing. If not specified, a default prompt will be used.
+        # Corresponds to the JSON property `customParsingPrompt`
+        # @return [String]
+        attr_accessor :custom_parsing_prompt
+      
+        # The maximum number of requests the job is allowed to make to the LLM model per
+        # minute. Consult https://cloud.google.com/vertex-ai/generative-ai/docs/quotas
+        # and your document size to set an appropriate value here. If unspecified, a
+        # default value of 5000 QPM would be used.
+        # Corresponds to the JSON property `maxParsingRequestsPerMin`
+        # @return [Fixnum]
+        attr_accessor :max_parsing_requests_per_min
+      
+        # The name of a LLM model used for parsing. Format: * `projects/`project_id`/
+        # locations/`location`/publishers/`publisher`/models/`model``
+        # Corresponds to the JSON property `modelName`
+        # @return [String]
+        attr_accessor :model_name
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @custom_parsing_prompt = args[:custom_parsing_prompt] if args.key?(:custom_parsing_prompt)
+          @max_parsing_requests_per_min = args[:max_parsing_requests_per_min] if args.key?(:max_parsing_requests_per_min)
+          @model_name = args[:model_name] if args.key?(:model_name)
+        end
+      end
+      
       # Specifies the transformation config for RagFiles.
       class GoogleCloudAiplatformV1beta1RagFileTransformationConfig
         include Google::Apis::Core::Hashable
@@ -26524,7 +27310,8 @@ module Google
       class GoogleCloudAiplatformV1beta1RagVectorDbConfig
         include Google::Apis::Core::Hashable
       
-        # The generic reusable api auth config.
+        # The generic reusable api auth config. Deprecated. Please use AuthConfig (
+        # google/cloud/aiplatform/master/auth.proto) instead.
         # Corresponds to the JSON property `apiAuth`
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1ApiAuth]
         attr_accessor :api_auth
@@ -27803,6 +28590,34 @@ module Google
         end
       end
       
+      # Retrieval config.
+      class GoogleCloudAiplatformV1beta1RetrievalConfig
+        include Google::Apis::Core::Hashable
+      
+        # The language code of the user.
+        # Corresponds to the JSON property `languageCode`
+        # @return [String]
+        attr_accessor :language_code
+      
+        # An object that represents a latitude/longitude pair. This is expressed as a
+        # pair of doubles to represent degrees latitude and degrees longitude. Unless
+        # specified otherwise, this object must conform to the WGS84 standard. Values
+        # must be within normalized ranges.
+        # Corresponds to the JSON property `latLng`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleTypeLatLng]
+        attr_accessor :lat_lng
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @language_code = args[:language_code] if args.key?(:language_code)
+          @lat_lng = args[:lat_lng] if args.key?(:lat_lng)
+        end
+      end
+      
       # Metadata related to retrieval in the grounding flow.
       class GoogleCloudAiplatformV1beta1RetrievalMetadata
         include Google::Apis::Core::Hashable
@@ -27930,6 +28745,59 @@ module Google
         end
       end
       
+      # Configuration for rolling deployments.
+      class GoogleCloudAiplatformV1beta1RolloutOptions
+        include Google::Apis::Core::Hashable
+      
+        # Percentage of allowed additional replicas. For autoscaling deployments, this
+        # refers to the target replica count.
+        # Corresponds to the JSON property `maxSurgePercentage`
+        # @return [Fixnum]
+        attr_accessor :max_surge_percentage
+      
+        # Absolute count of allowed additional replicas.
+        # Corresponds to the JSON property `maxSurgeReplicas`
+        # @return [Fixnum]
+        attr_accessor :max_surge_replicas
+      
+        # Percentage of replicas allowed to be unavailable. For autoscaling deployments,
+        # this refers to the target replica count.
+        # Corresponds to the JSON property `maxUnavailablePercentage`
+        # @return [Fixnum]
+        attr_accessor :max_unavailable_percentage
+      
+        # Absolute count of replicas allowed to be unavailable.
+        # Corresponds to the JSON property `maxUnavailableReplicas`
+        # @return [Fixnum]
+        attr_accessor :max_unavailable_replicas
+      
+        # ID of the DeployedModel that this deployment should replace.
+        # Corresponds to the JSON property `previousDeployedModel`
+        # @return [String]
+        attr_accessor :previous_deployed_model
+      
+        # Output only. Read-only. Revision number determines the relative priority of
+        # DeployedModels in the same rollout. The DeployedModel with the largest
+        # revision number specifies the intended state of the deployment.
+        # Corresponds to the JSON property `revisionNumber`
+        # @return [Fixnum]
+        attr_accessor :revision_number
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @max_surge_percentage = args[:max_surge_percentage] if args.key?(:max_surge_percentage)
+          @max_surge_replicas = args[:max_surge_replicas] if args.key?(:max_surge_replicas)
+          @max_unavailable_percentage = args[:max_unavailable_percentage] if args.key?(:max_unavailable_percentage)
+          @max_unavailable_replicas = args[:max_unavailable_replicas] if args.key?(:max_unavailable_replicas)
+          @previous_deployed_model = args[:previous_deployed_model] if args.key?(:previous_deployed_model)
+          @revision_number = args[:revision_number] if args.key?(:revision_number)
+        end
+      end
+      
       # Input for rouge metric.
       class GoogleCloudAiplatformV1beta1RougeInput
         include Google::Apis::Core::Hashable
@@ -32218,8 +33086,9 @@ module Google
       class GoogleCloudAiplatformV1beta1SchemaTextDataItem
         include Google::Apis::Core::Hashable
       
-        # Output only. Google Cloud Storage URI points to the original text in user's
-        # bucket. The text file is up to 10MB in size.
+        # Output only. Google Cloud Storage URI points to a copy of the original text in
+        # the Vertex-managed bucket in the user's project. The text file is up to 10MB
+        # in size.
         # Corresponds to the JSON property `gcsUri`
         # @return [String]
         attr_accessor :gcs_uri
@@ -36520,6 +37389,81 @@ module Google
         end
       end
       
+      # Configuration for Speculative Decoding.
+      class GoogleCloudAiplatformV1beta1SpeculativeDecodingSpec
+        include Google::Apis::Core::Hashable
+      
+        # Draft model speculation works by using the smaller model to generate candidate
+        # tokens for speculative decoding.
+        # Corresponds to the JSON property `draftModelSpeculation`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1SpeculativeDecodingSpecDraftModelSpeculation]
+        attr_accessor :draft_model_speculation
+      
+        # N-Gram speculation works by trying to find matching tokens in the previous
+        # prompt sequence and use those as speculation for generating new tokens.
+        # Corresponds to the JSON property `ngramSpeculation`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1SpeculativeDecodingSpecNgramSpeculation]
+        attr_accessor :ngram_speculation
+      
+        # The number of speculative tokens to generate at each step.
+        # Corresponds to the JSON property `speculativeTokenCount`
+        # @return [Fixnum]
+        attr_accessor :speculative_token_count
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @draft_model_speculation = args[:draft_model_speculation] if args.key?(:draft_model_speculation)
+          @ngram_speculation = args[:ngram_speculation] if args.key?(:ngram_speculation)
+          @speculative_token_count = args[:speculative_token_count] if args.key?(:speculative_token_count)
+        end
+      end
+      
+      # Draft model speculation works by using the smaller model to generate candidate
+      # tokens for speculative decoding.
+      class GoogleCloudAiplatformV1beta1SpeculativeDecodingSpecDraftModelSpeculation
+        include Google::Apis::Core::Hashable
+      
+        # Required. The resource name of the draft model.
+        # Corresponds to the JSON property `draftModel`
+        # @return [String]
+        attr_accessor :draft_model
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @draft_model = args[:draft_model] if args.key?(:draft_model)
+        end
+      end
+      
+      # N-Gram speculation works by trying to find matching tokens in the previous
+      # prompt sequence and use those as speculation for generating new tokens.
+      class GoogleCloudAiplatformV1beta1SpeculativeDecodingSpecNgramSpeculation
+        include Google::Apis::Core::Hashable
+      
+        # The number of last N input tokens used as ngram to search/match against the
+        # previous prompt sequence. This is equal to the N in N-Gram. The default value
+        # is 3 if not specified.
+        # Corresponds to the JSON property `ngramSize`
+        # @return [Fixnum]
+        attr_accessor :ngram_size
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @ngram_size = args[:ngram_size] if args.key?(:ngram_size)
+        end
+      end
+      
       # The speech generation config.
       class GoogleCloudAiplatformV1beta1SpeechConfig
         include Google::Apis::Core::Hashable
@@ -38245,6 +39189,12 @@ module Google
       class GoogleCloudAiplatformV1beta1SupervisedTuningDataStats
         include Google::Apis::Core::Hashable
       
+        # Output only. For each index in `truncated_example_indices`, the user-facing
+        # reason why the example was dropped. Must not include example itself.
+        # Corresponds to the JSON property `droppedExampleReasons`
+        # @return [Array<String>]
+        attr_accessor :dropped_example_reasons
+      
         # Output only. Number of billable characters in the tuning dataset.
         # Corresponds to the JSON property `totalBillableCharacterCount`
         # @return [Fixnum]
@@ -38255,7 +39205,9 @@ module Google
         # @return [Fixnum]
         attr_accessor :total_billable_token_count
       
-        # The number of examples in the dataset that have been truncated by any amount.
+        # Output only. The number of examples in the dataset that have been dropped. An
+        # example can be dropped for reasons including: too many tokens, contains an
+        # invalid image, contains too many images, etc.
         # Corresponds to the JSON property `totalTruncatedExampleCount`
         # @return [Fixnum]
         attr_accessor :total_truncated_example_count
@@ -38265,7 +39217,8 @@ module Google
         # @return [Fixnum]
         attr_accessor :total_tuning_character_count
       
-        # A partial sample of the indices (starting from 1) of the truncated examples.
+        # Output only. A partial sample of the indices (starting from 1) of the dropped
+        # examples.
         # Corresponds to the JSON property `truncatedExampleIndices`
         # @return [Array<Fixnum>]
         attr_accessor :truncated_example_indices
@@ -38306,6 +39259,7 @@ module Google
       
         # Update properties of this object
         def update!(**args)
+          @dropped_example_reasons = args[:dropped_example_reasons] if args.key?(:dropped_example_reasons)
           @total_billable_character_count = args[:total_billable_character_count] if args.key?(:total_billable_character_count)
           @total_billable_token_count = args[:total_billable_token_count] if args.key?(:total_billable_token_count)
           @total_truncated_example_count = args[:total_truncated_example_count] if args.key?(:total_truncated_example_count)
@@ -39426,6 +40380,11 @@ module Google
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1FunctionCallingConfig]
         attr_accessor :function_calling_config
       
+        # Retrieval config.
+        # Corresponds to the JSON property `retrievalConfig`
+        # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1RetrievalConfig]
+        attr_accessor :retrieval_config
+      
         def initialize(**args)
            update!(**args)
         end
@@ -39433,6 +40392,7 @@ module Google
         # Update properties of this object
         def update!(**args)
           @function_calling_config = args[:function_calling_config] if args.key?(:function_calling_config)
+          @retrieval_config = args[:retrieval_config] if args.key?(:retrieval_config)
         end
       end
       
@@ -40815,7 +41775,7 @@ module Google
         end
       end
       
-      # The Model Registry Model and Online Prediction Endpoint assiociated with this
+      # The Model Registry Model and Online Prediction Endpoint associated with this
       # TuningJob.
       class GoogleCloudAiplatformV1beta1TunedModel
         include Google::Apis::Core::Hashable
@@ -41005,7 +41965,7 @@ module Google
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1SupervisedTuningSpec]
         attr_accessor :supervised_tuning_spec
       
-        # The Model Registry Model and Online Prediction Endpoint assiociated with this
+        # The Model Registry Model and Online Prediction Endpoint associated with this
         # TuningJob.
         # Corresponds to the JSON property `tunedModel`
         # @return [Google::Apis::AiplatformV1beta1::GoogleCloudAiplatformV1beta1TunedModel]
@@ -43004,6 +43964,34 @@ module Google
         end
       end
       
+      # An object that represents a latitude/longitude pair. This is expressed as a
+      # pair of doubles to represent degrees latitude and degrees longitude. Unless
+      # specified otherwise, this object must conform to the WGS84 standard. Values
+      # must be within normalized ranges.
+      class GoogleTypeLatLng
+        include Google::Apis::Core::Hashable
+      
+        # The latitude in degrees. It must be in the range [-90.0, +90.0].
+        # Corresponds to the JSON property `latitude`
+        # @return [Float]
+        attr_accessor :latitude
+      
+        # The longitude in degrees. It must be in the range [-180.0, +180.0].
+        # Corresponds to the JSON property `longitude`
+        # @return [Float]
+        attr_accessor :longitude
+      
+        def initialize(**args)
+           update!(**args)
+        end
+      
+        # Update properties of this object
+        def update!(**args)
+          @latitude = args[:latitude] if args.key?(:latitude)
+          @longitude = args[:longitude] if args.key?(:longitude)
+        end
+      end
+      
       # Represents an amount of money with its currency type.
       class GoogleTypeMoney
         include Google::Apis::Core::Hashable