# Generated by API Evangelist (build-phrasing.py). Our phrasing, not observed demand.
overlay: 1.0.0
info:
  title: API Evangelist conversational phrasing for Microsoft Azure Run History Run Metrics API
  version: 1.0.0
extends: openapi/microsoft-azure-runmetrics-api-openapi.yml
actions:
- target: $.info
  update:
    x-apievangelist-phrasing:
      method: generated
      generated: '2026-09-26'
      generator: build-phrasing.py
      label: Generated by API Evangelist
      operations: 4
- target: $.paths['/history/v1.0/subscriptions/{subscriptionId}/resourceGroups/{resourceGroupName}/providers/Microsoft.MachineLearningServices/workspaces/{workspaceName}/experiments/{experimentName}/runs/{runId}/metrics'].post
  update:
    x-apievangelist-phrasing:
      intent: Log a metric to an experiment run
      effect: write
      questions:
      - How do I record a single metric, like accuracy, against an Azure ML run?
      - Can I log a table-style metric with cells and a schema to a run?
      instructions:
      - text: Log metric {name} to run {run_id} of experiment {experiment} in workspace {workspace}, group {resource_group}, subscription {subscription_id}.
        slots:
          name: requestBody.name
          run_id: path.runId
          experiment: path.experimentName
          workspace: path.workspaceName
          resource_group: path.resourceGroupName
          subscription_id: path.subscriptionId
      - text: Post one metric named {name} of type {metric_type} to run {run_id}.
        slots:
          name: requestBody.name
          metric_type: requestBody.metricType
          run_id: path.runId
      method: generated
      generated: '2026-09-26'
- target: $.paths['/history/v1.0/subscriptions/{subscriptionId}/resourceGroups/{resourceGroupName}/providers/Microsoft.MachineLearningServices/workspaces/{workspaceName}/experiments/{experimentName}/runs/{runId}/batch/metrics'].post
  update:
    x-apievangelist-phrasing:
      intent: Log a batch of metrics to an experiment run
      effect: write
      questions:
      - Can I send many metric values to a training run in a single call?
      - What is the batch way to upload metrics for an ML run?
      instructions:
      - text: Batch post metrics {values} to run {run_id} of experiment {experiment} in workspace {workspace}, group {resource_group}, subscription {subscription_id}.
        slots:
          values: requestBody.values
          run_id: path.runId
          experiment: path.experimentName
          workspace: path.workspaceName
          resource_group: path.resourceGroupName
          subscription_id: path.subscriptionId
      - text: 'Upload these metric values in bulk to run {run_id}: {values}.'
        slots:
          run_id: path.runId
          values: requestBody.values
      method: generated
      generated: '2026-09-26'
- target: $.paths['/history/v1.0/subscriptions/{subscriptionId}/resourceGroups/{resourceGroupName}/providers/Microsoft.MachineLearningServices/workspaces/{workspaceName}/experiments/{experimentName}/metrics/{metricId}'].get
  update:
    x-apievangelist-phrasing:
      intent: Get details of a logged metric
      effect: read
      questions:
      - How do I look up one logged metric by its metric ID in an experiment?
      - What details are stored for a specific run metric?
      instructions:
      - text: Get metric {metric_id} in experiment {experiment} of workspace {workspace}, group {resource_group}, subscription {subscription_id}.
        slots:
          metric_id: path.metricId
          experiment: path.experimentName
          workspace: path.workspaceName
          resource_group: path.resourceGroupName
          subscription_id: path.subscriptionId
      - text: Show the details of logged metric {metric_id}.
        slots:
          metric_id: path.metricId
      method: generated
      generated: '2026-09-26'
- target: $.paths['/history/v1.0/subscriptions/{subscriptionId}/resourceGroups/{resourceGroupName}/providers/Microsoft.MachineLearningServices/workspaces/{workspaceName}/experiments/{experimentName}/metrics:query'].post
  update:
    x-apievangelist-phrasing:
      intent: Query run metrics across an experiment
      effect: read
      questions:
      - How can I pull all run metrics for an experiment matching a filter?
      - Can I sort and page through experiment metrics with a continuation token?
      instructions:
      - text: Query run metrics in experiment {experiment} of workspace {workspace}, group {resource_group}, subscription {subscription_id}.
        slots:
          experiment: path.experimentName
          workspace: path.workspaceName
          resource_group: path.resourceGroupName
          subscription_id: path.subscriptionId
      - text: Find the top {top} metrics in experiment {experiment} matching {filter}, ordered by {order_by}.
        slots:
          top: requestBody.top
          experiment: path.experimentName
          filter: requestBody.filter
          order_by: requestBody.orderBy
      method: generated
      generated: '2026-09-26'