> For the complete documentation index, see [llms.txt](https://docs.cloudeka.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.cloudeka.ai/deka-gpu-mlops/runs/create-run.md).

# Create Run

Click the **+ Create Run** button. A window titled **Start a New Run** appears, containing several fields that need to be completed.

<figure><img src="/files/Jz9ey3uJQ9zGXJVNgc5c" alt=""><figcaption></figcaption></figure>

a. **Run Details**, select the pipeline that you want to use. You can either [create a new pipeline](/deka-gpu-mlops/pipelines/create-pipelines.md) or select an existing pipeline in the Cloudeka Deka MLOps Service Portal by clicking the **Use This Pipeline** button. After selecting the pipeline, choose the pipeline version and provide a name and description for the run. You must also specify the experiment that will be associated with the run. This information helps identify and organize your runs.

b. **Run Type**, this section allows you to specify the type of run to execute. Two options are available:

<figure><img src="/files/KxHJv2GbjGb9EqlOMDrk" alt=""><figcaption></figcaption></figure>

1. **One-off Run**: A one-off run is a single execution of the pipeline. The run is started manually and does not repeat automatically.
2. **Recurring Run**: A recurring run allows you to schedule the pipeline to run automatically. When selecting **Recurring Run**, configure the following settings:
   1. **Trigger Type**: You can choose one of the following trigger types:

      — **Periodic**: Runs the pipeline at regular intervals, such as every specified number of minutes, hours, days, weeks, or months.

      — **Cron**: Runs the pipeline based on a custom cron schedule.
   2. **Maximum Concurrent Runs**: Specifies the maximum number of pipeline runs that can be executed simultaneously.
   3. **Start Date and End Date (Optional)**: Allows you to specify the period during which the recurring run will be active.
   4. **Catch-up (Optional)**: If enabled, missed scheduled runs can be executed automatically when the system becomes available again.

c. **Pipeline Root**, specify the root directory for the pipeline. This location is used to store artifacts generated during the run. Defining a clear pipeline root helps manage outputs and logs efficiently.

d. **Run Parameters**, this section is used to define the parameters required by the pipeline. These parameters may include hyperparameters, input data paths, or other configuration values required for execution.

After completing all required fields, click the **Start** button to create the run.

<figure><img src="/files/ob4sA6ZV1KTyqDho7rpi" alt=""><figcaption></figcaption></figure>

If **Recurring Run** is selected as the run type, you will be redirected to the **Recurring Runs** page after the run is created successfully. This page displays information related to the recurring run that has been configured. To create additional recurring runs, refer to the [**Recurring Runs**](/deka-gpu-mlops/recurring-runs.md) documentation.

<figure><img src="/files/a2xWGAXVoS86VZFjJKoo" alt=""><figcaption></figcaption></figure>

If **One-off Run** is selected as the run type, you can click the **Start** button without configuring the run trigger settings. After the run is created successfully, you will be directed to the experiment details page.

<figure><img src="/files/BceX1nr6y7tXhYS2Wkpi" alt=""><figcaption></figcaption></figure>

After the process is completed successfully, you will be redirected to the **Active** submenu in the **Runs** menu.