Job Management

AI-Stack allows users to run code in a managed container using the Job feature. This is useful for scheduled tasks, batch processing, or long-running ML training jobs.

Create a Job

To launch a job:

1. Click the ➕ icon in the top-left corner of Job List

2. Enter the Job Name

3. Select:

   • GPU Type and Quantity

   • CPU and RAM Specs

   • Container Image

4. Enter the command to execute:

echo "Hello World"

1. Click [Next Step | Advanced Settings] (optional)

2. Review and click [Create]

⏳ Job status will update from Job Initialize → Succeeded upon completion.

---

Job Scheduling

You can schedule jobs to run automatically using one of two methods:

Method 1: From Job Scheduling Page

1

Go to Job Management → Job Scheduling

2

Click ➕ to Create Schedule

3

Set:

• Schedule Name and Description

• Trigger Time (e.g., daily at 20:10)

• Start/End Time

• Timezone

4

Click [Add Job] and select a pre-created Job Template

5

Configure whether to [Enable Concurrency (Multi-Threading)].

a. 🔁 Multi-threaded: Jobs run in parallel

b. 🔄 Single-threaded: Each job runs one at a time

6

Review and click [Create Schedule]

📅 Upcoming executions will show in both Job List and Job Scheduling.

---

Method 2: From Job Templates Page

1

Go to Advanced Settings → Job Templates

2

Select a template and click [Run Job with Schedule]

3

Configure:

• Start/End Time

• Trigger Frequency

• Timezone

• Concurrency

4

Click [Create]

🧠 Perfect for recurring training tasks or inference jobs.

Deleting Job /Scheduled Jobs

Jobs created via a schedule cannot be deleted individually.

To remove them:

• Delete the entire schedule in Job Scheduling

• All jobs under that schedule will be removed

💡 Pro Tip Use Job Templates + Scheduling to automate regular tasks — like retraining models, generating reports, or monitoring data pipelines!