Logs Section

The Logs section provides detailed insights into the execution of jobs, tracking their progress, errors, and relevant system information. This section describes logs for both Models section as well as Scenarios section.

Accessing the Logs

Location: The logs can be accessed by navigating to the Logs tab within the job execution window.
Log Display: The Logs section presents an overview of all runs. Each row corresponds to a job execution, displaying the following fields:
- Job Identifier: A unique identifier for each job execution.
- Progress: Indicates the completion status, such as "1/1", meaning the job has completed.
- Type: The type of scenario executed (e.g., Binary Classification, Recommendation...).
- Elapsed Time: Shows how long the job took to execute.
- Start Date: The date and time the job began execution.
If errors occur during a job execution, they will be highlighted with a red icon next to the job identifier.

Viewing Individual Log Details

To view the details of a specific job log:

Click on the three vertical dots (More options) next to the job entry.
Select View logs from the dropdown.

This opens the full log for that specific job, providing detailed information about its execution. The logs typically include:

INFO statements: General system information, such as timestamped messages about the start of various operations or warnings.
Traceback (in case of errors): Detailed Python traceback logs in case an error occurred during job execution. The logs show:
- The file and line where the error occurred.
- A description of the error (e.g., FileNotFoundError: [Errno 2] No such file or directory).
This is useful for developers and administrators to debug issues, especially when a job fails or behaves unexpectedly.

Downloading Logs

For in-depth analysis, the logs can be downloaded:

From the More options menu next to each job, select Download logs.
This will save the full log file to your local system, where it can be reviewed in any text editor or log analysis tool.