experiment_runs

Creates, updates, deletes, gets or lists an experiment_runs resource.

Overview

Name	experiment_runs
Type	Resource
Id	databricks_workspace.machinelearning.experiment_runs

Fields

The following fields are returned by SELECT queries:

searchruns

Name	Datatype	Description
data	object
info	object
inputs	object

getrun

Name	Datatype	Description
data	object
info	object
inputs	object

Methods

The following methods are available for this resource:

Name	Accessible by	Required Params	Optional Params	Description
searchruns	select	deployment_name		Searches for runs that satisfy expressions.
getrun	select	deployment_name		Gets the metadata, metrics, params, and tags for a run. In the case where multiple metrics with the same key are logged for a run, return only the value with the latest timestamp.
createrun	insert	deployment_name		Creates a new run within an experiment. A run is usually a single execution of a machine learning or data ETL pipeline. MLflow uses runs to track the
updaterun	update	deployment_name		Updates run metadata.
deleteruns	delete	deployment_name		Bulk delete runs in an experiment that were created prior to or at the specified timestamp. Deletes at most max_runs per request. To call this API from a Databricks Notebook in Python, you can use the client code snippet on
deleterun	delete	deployment_name		Marks a run for deletion.
logbatch	exec	deployment_name		Logs a batch of metrics, params, and tags for a run. If any data failed to be persisted, the server will respond with an error (non-200 status code).
loginputs	exec	deployment_name		Experimental: This API may change or be removed in a future release without warning.
logmetric	exec	deployment_name		Logs a metric for a run. A metric is a key-value pair (string key, float value) with an associated timestamp. Examples include the various metrics that represent ML model accuracy. A metric can be logged multiple times.
logmodel	exec	deployment_name		Experimental: This API may change or be removed in a future release without warning.
logparam	exec	deployment_name		Logs a param used for a run. A param is a key-value pair (string key, string value). Examples include hyperparameters used for ML model training and constant dates and values used in an ETL pipeline. A param can be logged only once for a run.
restorerun	exec	deployment_name		Restores a deleted run.
restoreruns	exec	deployment_name		Bulk restore runs in an experiment that were deleted no earlier than the specified timestamp. Restores at most max_runs per request. To call this API from a Databricks Notebook in Python, you can use the client code snippet on

Parameters

Parameters can be passed in the WHERE clause of a query. Check the Methods section to see which parameters are required or optional for each operation.

Name	Datatype	Description
deployment_name	string	The Databricks Workspace Deployment Name (default: dbc-abcd0123-a1bc)

SELECT examples

searchruns

Searches for runs that satisfy expressions.

SELECT
data,
info,
inputs
FROM databricks_workspace.machinelearning.experiment_runs
WHERE deployment_name = '{{ deployment_name }}' -- required;

getrun

Gets the metadata, metrics, params, and tags for a run. In the case where multiple metrics with the same key are logged for a run, return only the value with the latest timestamp.

SELECT
data,
info,
inputs
FROM databricks_workspace.machinelearning.experiment_runs
WHERE deployment_name = '{{ deployment_name }}' -- required;

INSERT examples

createrun

Creates a new run within an experiment. A run is usually a single execution of a machine learning or data ETL pipeline. MLflow uses runs to track the

INSERT INTO databricks_workspace.machinelearning.experiment_runs (
data__experiment_id,
data__user_id,
data__start_time,
data__tags,
deployment_name
)
SELECT 
'{{ experiment_id }}',
'{{ user_id }}',
{{ start_time }},
'{{ tags }}',
'{{ deployment_name }}'
RETURNING
run
;

Manifest

# Description fields are for documentation purposes
- name: experiment_runs
  props:
    - name: deployment_name
      value: string
      description: Required parameter for the experiment_runs resource.
    - name: experiment_id
      value: string
    - name: user_id
      value: string
    - name: start_time
      value: integer
    - name: tags
      value: Array of object

UPDATE examples

updaterun

Updates run metadata.

UPDATE databricks_workspace.machinelearning.experiment_runs
SET 
data__run_id = '{{ run_id }}',
data__run_uuid = '{{ run_uuid }}',
data__status = '{{ status }}',
data__end_time = {{ end_time }}
WHERE 
deployment_name = '{{ deployment_name }}' --required
RETURNING
run_info;

DELETE examples

deleteruns

Bulk delete runs in an experiment that were created prior to or at the specified timestamp. Deletes at most max_runs per request. To call this API from a Databricks Notebook in Python, you can use the client code snippet on

DELETE FROM databricks_workspace.machinelearning.experiment_runs
WHERE deployment_name = '{{ deployment_name }}' --required;

deleterun

Marks a run for deletion.

DELETE FROM databricks_workspace.machinelearning.experiment_runs
WHERE deployment_name = '{{ deployment_name }}' --required;

Lifecycle Methods

logbatch

Logs a batch of metrics, params, and tags for a run. If any data failed to be persisted, the server will respond with an error (non-200 status code).

EXEC databricks_workspace.machinelearning.experiment_runs.logbatch 
@deployment_name='{{ deployment_name }}' --required 
@@json=
'{
"run_id": "{{ run_id }}", 
"metrics": "{{ metrics }}", 
"params": "{{ params }}", 
"tags": "{{ tags }}"
}';

loginputs

Experimental: This API may change or be removed in a future release without warning.

EXEC databricks_workspace.machinelearning.experiment_runs.loginputs 
@deployment_name='{{ deployment_name }}' --required 
@@json=
'{
"run_id": "{{ run_id }}", 
"datasets": "{{ datasets }}"
}';

logmetric

Logs a metric for a run. A metric is a key-value pair (string key, float value) with an associated timestamp. Examples include the various metrics that represent ML model accuracy. A metric can be logged multiple times.

EXEC databricks_workspace.machinelearning.experiment_runs.logmetric 
@deployment_name='{{ deployment_name }}' --required 
@@json=
'{
"run_id": "{{ run_id }}", 
"run_uuid": "{{ run_uuid }}", 
"key": "{{ key }}", 
"value": "{{ value }}", 
"timestamp": "{{ timestamp }}", 
"step": "{{ step }}"
}';

logmodel

Experimental: This API may change or be removed in a future release without warning.

EXEC databricks_workspace.machinelearning.experiment_runs.logmodel 
@deployment_name='{{ deployment_name }}' --required 
@@json=
'{
"run_id": "{{ run_id }}", 
"model_json": "{{ model_json }}"
}';

logparam

Logs a param used for a run. A param is a key-value pair (string key, string value). Examples include hyperparameters used for ML model training and constant dates and values used in an ETL pipeline. A param can be logged only once for a run.

EXEC databricks_workspace.machinelearning.experiment_runs.logparam 
@deployment_name='{{ deployment_name }}' --required 
@@json=
'{
"run_id": "{{ run_id }}", 
"run_uuid": "{{ run_uuid }}", 
"key": "{{ key }}", 
"value": "{{ value }}"
}';

restorerun

Restores a deleted run.

EXEC databricks_workspace.machinelearning.experiment_runs.restorerun 
@deployment_name='{{ deployment_name }}' --required 
@@json=
'{
"run_id": "{{ run_id }}"
}';

restoreruns

Bulk restore runs in an experiment that were deleted no earlier than the specified timestamp. Restores at most max_runs per request. To call this API from a Databricks Notebook in Python, you can use the client code snippet on

EXEC databricks_workspace.machinelearning.experiment_runs.restoreruns 
@deployment_name='{{ deployment_name }}' --required 
@@json=
'{
"experiment_id": "{{ experiment_id }}", 
"min_timestamp_millis": "{{ min_timestamp_millis }}", 
"max_runs": "{{ max_runs }}"
}';