Types

MLFlow

Base type which defines location and version for MLFlow API service.

Fields

• apiroot::String: API root URL, e.g. http://localhost:5000/api
• apiversion::Union{Integer, AbstractFloat}: used API version, e.g. 2.0
• headers::AbstractDict: HTTP headers to be provided with the REST API requests.
• username::Union{Nothing, String}: username for basic authentication.
• password::Union{Nothing, String}: password for basic authentication.

Examples

mlf = MLFlow()

remote_url="https://<your-server>.cloud.databricks.com"; # address of your remote server
mlf = MLFlow(remote_url, headers=Dict("Authorization" => "Bearer <your-secret-token>"))

Tag <: LoggingData

Generic tag type for MLFlow entities.

Fields

• key::String: The tag key.
• value::String: The tag value.

Dataset

Represents a reference to data used for training, testing, or evaluation during the model development process.

Fields

• name::String: The name of the dataset.
• digest::String: The digest of the dataset.
• source_type::String: The type of the dataset source.
• source::String: Source information for the dataset.
• schema::String: The schema of the dataset. This field is optional.
• profile::String: The profile of the dataset. This field is optional.

DatasetInput

Represents a dataset and input tags.

Fields

• tags::Array{Tag}: A list of tags for the dataset input.
• dataset::Dataset: The dataset being used as a run input.

FileInfo

Fields

• path::String: Path relative to the root artifact directory run.
• is_dir::Bool: Whether the path is a directory.
• file_size::Int64: Size in bytes. Unset for directories.

ModelInput

Represents a logged model or RegisteredModel version input to a Run.

Fields

• model_id::String: The unique identifier of the model.

ModelOutput

Represents a logged model output of a Run.

Fields

• model_id::String: The unique identifier of the model.
• step::Int64: Step at which the model was produced.

ModelMetric

Metric associated with a model, represented as a key-value pair.

Fields

• key::String: Key identifying this metric.
• value::Float64: Value associated with this metric.
• timestamp::Int64: The timestamp at which this metric was recorded.
• step::Union{Int64, Nothing}: Step at which to log the metric.

ModelParam

Param for a model version.

Fields

• name::String: Name of the param.
• value::String: Value of the param associated with the name

ModelVersionDeploymentJobState

Deployment job state for a model version.

Fields

• job_id::String: The job ID.
• run_id::String: The run ID.
• job_state::State.StateEnum: The state of the job.
• run_state::DeploymentJobRunState.DeploymentJobRunStateEnum: The state of the run.
• current_task_name::String: The current task name.

ModelVersion

Fields

• name::String: Unique name of the model.
• version::String: Model's version number.
• creation_timestamp::Int64: Timestamp recorded when this model_version was created.
• last_updated_timestamp::Int64: Timestamp recorded when metadata for this model_version was last updated.
• user_id::Union{String, Nothing}: User that created this model_version.
• current_stage::String: Current stage for this model_version.
• description::String: Description of this model_version.
• source::String: URI indicating the location of the source model artifacts, used when creating model_version.
• run_id::String: MLflow run ID used when creating model_version, if source was generated by an experiment run stored in MLflow tracking server.
• status::ModelVersionStatusEnum: Current status of model_version.
• status_message::String: Details on current status, if it is pending or failed.
• tags::Array{Tag}: Additional metadata key-value pairs.
• run_link::Union{String, Nothing}: Direct link to the run that generated this version.
• aliases::Array{String}: Aliases pointing to this model_version.
• model_id::Union{String, Nothing}: Optional model_id for ModelVersion.
• model_params::Array{ModelParam}: Optional parameters for the model.
• model_metrics::Array{ModelMetric}: Optional metrics for the model.
• deployment_job_state::Union{ModelVersionDeploymentJobState, Nothing}: Deployment job state.

RegisteredModel

Fields

• name::String: Unique name for the model.
• creation_timestamp::Int64: Timestamp recorded when this RegisteredModel was created.
• last_updated_timestamp::Int64: Timestamp recorded when metadata for this RegisteredModel was last updated.
• user_id::Union{String, Nothing}: User that created this RegisteredModel.
• description::Union{String, Nothing}: Description of this RegisteredModel.
• latest_versions::Array{ModelVersion}: Collection of latest model versions for each stage. Only contains models with current READY status.
• tags::Array{Tag}: Additional metadata key-value pairs.
• aliases::Array{RegisteredModelAlias}: Aliases pointing to model versions associated with this RegisteredModel.
• deployment_job_id::String: Deployment job id for this model.
• deployment_job_state::State: Deployment job state for this model.

RegisteredModelAlias

Alias for a registered model.

Fields

• alias::String: The name of the alias.
• version::String: The model version number that the alias points to.

Experiment

Fields

• experiment_id::Integer: Unique identifier for the experiment.
• name::String: Human readable name that identifies the experiment.
• artifact_location::String: Location where artifacts for the experiment are stored.
• lifecycle_stage::String: Current life cycle stage of the experiment: “active” or “deleted”. Deleted experiments are not returned by APIs.
• last_update_time::Int64: Last update time.
• creation_time::Int64: Creation time.
• tags::Array{Tag}: Additional metadata key-value pairs.

Run

A single run.

Fields

• info::RunInfo: Metadata of the run.
• data::RunData: Run data (metrics, params, and tags).
• inputs::RunInputs: Run inputs.
• outputs::RunOutputs: Run outputs.

Param <: LoggingData

Param associated with a run.

Fields

• key::String: Key identifying this param.
• value::String: Value associated with this param.

Metric <: LoggingData

Metric associated with a run, represented as a key-value pair.

Fields

• key::String: Key identifying this metric.
• value::Float64: Value associated with this metric.
• timestamp::Int64: The timestamp at which this metric was recorded.
• step::Union{Int64, Nothing}: Step at which to log the metric.

RunInputs

Run data (metrics, params, and tags).

Fields

• metrics::Array{Metric}: Run metrics.
• params::Array{Param}: Run parameters.
• tags::Array{Tag}: Additional metadata key-value pairs.

RunInfo

Metadata of a single run.

Fields

• run_id::String: Unique identifier for the run.
• run_name::String: The name of the run.
• experiment_id::String: The experiment ID.
• status::RunStatus: Current status of the run.
• start_time::Int64: Unix timestamp of when the run started in milliseconds.
• end_time::Int64: Unix timestamp of when the run ended in milliseconds.
• artifact_uri::String: URI of the directory where artifacts should be uploaded. This can be a local path (starting with “/”), or a distributed file system (DFS) path, like s3://bucket/directory or dbfs:/my/directory. If not set, the local ./mlruns directory is chosen.
• lifecycle_stage::String: Current life cycle stage of the experiment: "active" or "deleted".

RunInputs

Run inputs.

Fields

• dataset_inputs::Array{DatasetInput}: Dataset inputs to the Run.

RunOutputs

Outputs of a Run.

Fields

• model_outputs::Array{ModelOutput}: Model outputs of the Run.

User

Fields

• id::String: User ID.
• username::String: Username.
• is_admin::Bool: Whether the user is an admin.
• experiment_permissions::Array{ExperimentPermission}: All experiment permissions explicitly granted to the user.
• registered_model_permissions::Array{RegisteredModelPermission}: All registered model explicitly granted to the user.

ExperimentPermission

Fields

• experiment_id::String: Experiment id.
• user_id::String: User id.
• permission::Permission.PermissionEnum: Permission granted.

RegisteredModelPermission

Fields

• name::String: RegisteredModel name.
• user_id::String: User id.
• permission::Permission.PermissionEnum: Permission granted.

Scorer

Represents a scorer for MLFlow experiments.

Fields

• experiment_id: The experiment ID.
• name: The scorer name.
• version: Version number of the scorer.
• scorer_id: Unique identifier for the scorer.
• serialized_scorer: The serialized scorer string (JSON).
• creation_time: Creation time in milliseconds since epoch.

Webhook

Represents a webhook configuration for MLflow events.

Fields

• webhook_id::String: Unique identifier for the webhook.
• name::String: Name of the webhook.
• description::Union{String, Nothing}: Description of the webhook.
• url::String: URL to send webhook events to.
• events::Array{WebhookEvent}: Events this webhook is subscribed to.
• status::WebhookStatus.WebhookStatusEnum: Current status of the webhook.
• creation_timestamp::Int64: Timestamp when webhook was created.
• last_updated_timestamp::Int64: Timestamp when webhook was last updated.

WebhookTestResult

Result of a webhook test.

Fields

• success::Bool: Whether the test succeeded.
• response_status::Int: HTTP response status code.
• response_body::String: HTTP response body.

GatewaySecretInfo

Represents metadata about a gateway secret (does not include encrypted value).

Fields

• secret_id: Unique identifier for the secret.
• secret_name: User-friendly name for the secret.
• provider: Optional LLM provider name.
• created_by: Username of the creator.
• last_updated_by: Username of the last updater.
• created_at: Creation time in milliseconds since epoch.
• last_updated_at: Last update time in milliseconds since epoch.

GatewayModelDefinition

Represents a reusable gateway model definition.

Fields

• model_definition_id: Unique identifier for the model definition.
• name: User-friendly name for the model definition.
• secret_id: ID of the secret containing authentication credentials.
• secret_name: Name of the secret for display purposes.
• provider: LLM provider name.
• model_name: Provider-specific model identifier.
• created_by: Username of the creator.
• last_updated_by: Username of the last updater.
• created_at: Creation time in milliseconds since epoch.
• last_updated_at: Last update time in milliseconds since epoch.

GatewayEndpoint

Represents a gateway endpoint.

Fields

• endpoint_id: Unique identifier for the endpoint.
• name: Name of the endpoint.
• model_mappings: Array of model mappings bound to this endpoint.
• routing_strategy: Routing strategy for the endpoint.
• fallback_config: Fallback configuration.
• created_by: Username of the creator.
• last_updated_by: Username of the last updater.
• created_at: Creation time in milliseconds since epoch.
• last_updated_at: Last update time in milliseconds since epoch.
• tags: Array of endpoint tags.
• experiment_id: ID of the MLflow experiment where traces are logged.
• usage_tracking: Whether usage tracking is enabled.

GatewayEndpointConfig

Represents configuration for a gateway endpoint.

Fields

• model_definition_id: ID of the model definition.
• route: Route configuration as JSON string.
• limits: Rate limits as JSON string.
• auth: Auth configuration as JSON string.
• metadata: Additional metadata as JSON string.

GatewayEndpointBinding

Represents a binding between a gateway endpoint and an MLflow resource.

Fields

• endpoint_id: ID of the endpoint.
• resource_type: Type of MLflow resource (e.g., "scorer").
• resource_id: ID of the MLflow resource.
• created_by: Username of the creator.
• created_at: Creation time in milliseconds since epoch.

GatewayEndpointModelConfig

Configuration for a model attached to an endpoint.

Fields

• model_definition_id: ID of the model definition.
• linkage_type: Type of linkage.
• weight: Routing weight for traffic distribution.
• fallback_order: Order for fallback attempts.

GatewayEndpointModelMapping

Mapping between an endpoint and a model definition.

Fields

• mapping_id: Unique identifier for this mapping.
• endpoint_id: ID of the endpoint.
• model_definition_id: ID of the model definition.
• model_definition: The full model definition (populated via JOIN).
• weight: Routing weight for traffic distribution.
• created_at: Timestamp when the mapping was created.
• created_by: User ID who created the mapping.
• linkage_type: Type of linkage.
• fallback_order: Order for fallback attempts.

GatewayEndpointTag

Tag associated with an endpoint.

Fields

• key: Tag key.
• value: Tag value.

GatewayBudgetWindow

Represents a budget window for rate limiting.

Fields

• budget_policy_id: ID of the budget policy.
• window_start_ms: Window start timestamp.
• window_end_ms: Window end timestamp.
• current_spend: Current spend in the window.

GatewayBudgetPolicy

A budget policy for gateway endpoints.

Fields

• budget_policy_id: Unique identifier for the budget policy.
• budget_unit: Budget measurement unit.
• budget_amount: Budget amount.
• duration: Budget duration configuration.
• target_scope: Target scope for the budget policy.
• budget_action: Action to take when budget is exceeded.
• created_by: Username of the creator.
• last_updated_by: Username of the last updater.
• created_at: Creation time in milliseconds since epoch.
• last_updated_at: Last update time in milliseconds since epoch.

BudgetDuration

Fixed window duration: a (unit, value) pair that defines how long a budget window is.

Fields

• unit: Unit of time (MINUTES, HOURS, DAYS, WEEKS, MONTHS).
• value: Number of units per window.

FallbackConfig

Configuration for fallback routing.

Fields

• strategy: The fallback strategy.
• max_attempts: The max attempts for fallback routing.

MultipartUploadCredential

Credentials for a multipart upload part.

Fields

• part_number::Int64: Part number (1-indexed).
• upload_url::String: Presigned URL for uploading this part.
• headers::Dict{String,String}: Required headers for the upload request.

MultipartUploadPart

A part in a multipart upload.

Fields

• part_number::Int64: Part number (1-indexed).
• etag::String: ETag of the uploaded part.

PromptOptimizationJob

Represents a prompt optimization job entity.

Fields

• job_id: Unique identifier for the optimization job.
• run_id: MLflow run ID where optimization metrics are stored.
• state: Current state of the job.
• experiment_id: ID of the MLflow experiment.
• source_prompt_uri: URI of the source prompt.
• optimized_prompt_uri: URI of the optimized prompt (only set if completed).
• config: Configuration for the optimization job.
• creation_timestamp_ms: Creation timestamp in milliseconds.
• completion_timestamp_ms: Completion timestamp in milliseconds.
• tags: Tags associated with the job.
• initial_eval_scores: Initial evaluation scores before optimization.
• final_eval_scores: Final evaluation scores after optimization.

PromptOptimizationJobConfig

Configuration for a prompt optimization job.

Fields

• optimizer_type: The optimizer type to use.
• dataset_id: ID of the EvaluationDataset containing training data.
• scorers: List of scorer names.
• optimizer_config_json: JSON-serialized optimizer-specific configuration.

PromptOptimizationJobTag

Represents a tag for a prompt optimization job.

Fields

• key: Tag key.
• value: Tag value.

InitialEvalScoresEntry

Represents an initial evaluation scores entry.

Fields

• scorer_name: Name of the scorer.
• score: Score value.

FinalEvalScoresEntry

Represents a final evaluation scores entry.

Fields

• scorer_name: Name of the scorer.
• score: Score value.

JobStateInfo

Represents the state of a job.

Fields

• state: Current state (PENDING, RUNNING, COMPLETED, FAILED, CANCELED).
• message: Optional error or status message.