Skip to main content

The Experiments Table

The experiments table is a customizable list of experiments associated with a project. From the experiments table, view experiment details, and work with experiments (reset, clone, enqueue, create tracking leaderboards to monitor experimentation, and more). The experiments table's auto-refresh allows users to continually monitor experiment progress.


To assist in focusing on active experimentation, experiments and models can be archived, so they will not appear in the active experiments and models tables. See Archiving.

Experiment table

Experiments table columns#

The experiments table default and customizable columns are described in the following table.

TYPEType of experiment. ClearML supports multiple task types for experimentation, and a variety of workflows and use cases.Default
NAMEExperiment name.Default
TAGSDescriptive, user-defined, color-coded tags assigned to experiments. Use tags to classify experiments, and filter the list. See tagging experiments.Default
STATUSExperiment state (status). See a list of the task states and state transtions.Default
PROJECTName of experiment's project.Default
USERUser who created or cloned the experiment.Default (hidden)
STARTEDElapsed time since the experiment started. To view the date and time of start, hover over the elapsed time.Default
UPDATEDElapsed time since the last update to the experiment. To view the date and time of update, hover over the elapsed time.Default
ITERATIONLast or most recent iteration of the experiment.Default
DESCRIPTIONA description of the experiment. For cloned experiments, the description indicates it was auto generated with a timestamp.Default (hidden)
RUN TIMEThe current / total running time of the experiment.Default (hidden)
MetricsAdd metrics column (last, minimum, and / or maximum values). The metrics depend upon the experiments in the table. See adding metrics.Customizable
HyperparametersAdd hyperparameters. The hyperparameters depend upon the experiments in the table. See adding hyperparameters.Customizable

Customizing the experiments table#

The experiments table can be customized by:

  • Showing / hiding default columns
  • Adding metrics and hyperparameters
  • Sorting
  • Filtering

Use experiments table customization for various use cases, including:

  • Creating a leaderboard that will update in real time with experiment performance, which can be shared and stored.
  • Sorting models by metrics - Models are associated with the experiments that created them. For each metric, use the last value, the minimal value, and / or the maximal value.
  • Tracking hyperparameters - Track hyperparameters by adding them as columns, and applying filters and sorting.

Changes are persistent (cached in the browser), and represented in the URL so customized settings can be saved in a browser bookmark and shared with other ClearML users to collaborate.


The following experiments-table customizations are saved on a per project basis:

  • Columns order
  • Column width
  • Active sort order
  • Active filters
  • Custom columns

If a project has sub-projects, the experiments can be viewed by their sub-project groupings or together with all the experiments in the project. The customizations of these two views are saved separately.

Adding metrics and / or hyperparameters#

Experiment table customization gif

Add metrics and / or hyperparameters columns to the experiments table. The metrics and hyperparameters depend upon the experiments in the table.

To add metrics:#

  • Click Setting Gear > + METRICS > Expand a metric > Select the LAST (value), MIN (minimal value), and / or MAX (maximal value) checkboxes.

To add hyperparameters:#

  • Click Setting Gear > + HYPER PARAMETERS > Expand a section > Select the hyperparameter checkboxes.

Filtering experiments#

Filtering table gif

Filters can be applied by clicking Filter on a column, and the relevant filter appears.

There are a few types of filters:

  • Value set - Choose which values to include from a list of all values in the column
  • Numerical ranges - Insert minimum and / or maximum value
  • Date ranges - Insert starting and / or ending date and time
  • Tags - Choose which tags to include from a list of all tags used in the column. Additionally, tags can be filtered using the ANY or ALL options, which correspond to the logical "AND" and "OR" respectively. These options appear on the top of the tag list.

Once a filter is applied to a column, its filter icon will appear with a highlighted dot on its top right (Filter on ).

Using other customization features#

To use other customization features:

  • Show / hide columns - Click Setting Gear > select or clear the checkboxes of columns to show or hide.
  • Sort columns - According to metrics and hyperparameters, type of experiment, experiment name, start and last update elapsed time, and last iteration.
  • Dynamic column ordering - Drag a column title to a different position.
  • Column resizing - In the column heading, drag to a new size.
  • Column autofit - In the column heading, double click a column separator.

Experiment actions#

The following table describes the actions that can be done from the experiments table, including the states that allow each operation.

ActionDescriptionStates Valid for the ActionState Transition
View detailsView experiment details in the experiments table, the info panel (keep the experiments table in view), or the full screen details view.Any stateNone
Manage a queueIf an experiment is Pending in a queue, view the utilization of that queue, manage that queue (remove experiments and change the order of experiments), and view information about the worker(s) listening to the queue. See the Workers and queues page.EnqueuedNone
View a workerIf an experiment is Running, view resource utilization, worker details, and queues to which a worker is listening.RunningNone
ShareFor ClearML Hosted Service users only, share an experiment and its model with a ClearML Hosted Service user in another workspace.Any stateNone
ArchiveTo more easily work with active experiments, move an experiment to the archive. See Archiving.Any stateNone
EnqueueAdd an experiment to a queue for a worker or workers (listening to the queue) to execute.DraftPending
DequeueRemove an experiment from a queue.PendingDraft
ResetDelete the log and output from a previous run of an experiment (for example, before rerunning it).Completed, Aborted, or FailedDraft
AbortManually terminate a Running experiment.RunningAborted
PublishPublish an experiment to prevent changes to its tracking data, inputs, and outputs. Published experiments and their models are read-only. Published experiments cannot be enqueued, but they can be cloned, and their clones can be edited, tuned, and enqueued.Completed, Aborted, or Failed.Published
TagsTag experiments with color-coded labels to assist you in organizing your work. See tagging experiments.Any stateNone
CloneMake an exact, editable copy of an experiment (for example, to reproduce an experiment, but keep the original).DraftNewly Cloned Experiment is Draft
MoveMove an experiment to another project.Any stateNone

These actions can be accessed with the context menu (when right-clicking an experiment or clicking the menu button Menu in an experiment's info panel).

Most of the actions mentioned in the chart above can be performed on multiple experiments at once. Select multiple experiments, then use either the context menu, or the bar that appears at the bottom of the page, to perform operations on the selected experiments. Actions can be performed only on the experiments that match the action criteria (for example, only Running experiments can be aborted). The context menu shows the number of experiments that can be affected by each action. The same information can be found in the bottom menu, in a tooltip that appears when hovering over an action icon.

Experiment table batch operations

Creating an experiment leaderboard#

Filter & sort the experiments of any project to create a leaderboard that can be shared and stored. This leaderboard updates in real time with experiment performance and outputs.

Modify the experiment table in the following ways to create a customized leaderboard:

  • Add experiment configuration (hyperparameters)
  • Edit and add experiments properties
  • Add reported metrics, any time series reported metric can be selected, then select the last reported value, or the minimum / maximum reported value.
  • Filter based on user (dropdown and select) or experiment types
  • Add specific tags and filter based on them.

Now the table can be sorted based on any of the columns (probably one of the performance metrics). Select to filter experiments based on their name by using the search bar.

The final dashboard can be shared by copying the URL from the address bar, this address will replicate the exact same dashboard on any browser. The dashboard can also be bookmarked for later use.

Experiment table sharing