Simplify Your Work with Run AI Workspaces

Table of Contents

1. Introduction
2. Overview of Run AI Workspaces
3. Benefits of Run AI Workspaces
4. How to Access Run AI Workspaces
5. Creating a Workspace
   • Selecting Tools and Versions
   • Choosing Compute Resources
   • Selecting Data Sources
   • Creating the Workspace
6. Connecting to a Workspace
   • Accessing Workspace Tools
   • Working with Jupiter Notebook
   • Using MLflow
7. Managing Workspaces
   • Starting and Stopping Workspaces
   • Sharing Compute Resources
8. Customizing Workspaces
   • Creating Templates
   • Editing Environment Settings
   • Modifying Compute Resources
9. Conclusion

Introduction

Welcome to this demo video of the new Run AI Workspaces. This video provides an overview of the capabilities and features of Run AI Workspaces, a platform that offers a Simplified and efficient way for data scientists to Create and manage their work environments.

Overview of Run AI Workspaces

Run AI Workspaces is designed to enable data scientists to focus on their research without the hassle of building container images or configuring networking and storage. With Run AI Workspaces, data scientists can easily create and connect to their workspaces, choosing the tools and versions they need.

Benefits of Run AI Workspaces

• Simplified workspace creation process
• Easy access to various tools and versions
• Efficient resource utilization with GPU capabilities
• Seamless integration with Git repositories and data sources
• Flexibility to start and stop workspaces as needed

How to Access Run AI Workspaces

To access Run AI Workspaces, log in to your Run AI account and navigate to the Workspaces page. Here, you will find all the workspaces assigned to your project, including active and stopped workspaces. From the Workspaces page, you can choose the workspace you want to use and connect to it.

Creating a Workspace

To create a new workspace, go to the New Workspaces page and select your project. You have the option to choose an existing template that defines the environment, data source, and compute resources, or create a new one from scratch. Enter the required details, such as the desired tools, versions, and compute resources, and select the data sources you need. Once done, click on "Create Workspace."

Selecting Tools and Versions

When creating a workspace, You can choose from a variety of pre-defined environments that have been set up by the administrators. For example, if you want to use TensorFlow version 2.4 with Jupiter Notebook, simply select the appropriate environment.

Choosing Compute Resources

Based on your requirements, you can select the amount of resources you need for your workspace. If you don't require any GPUs, you can opt for a CPU-only resource. Alternatively, you can choose a fraction of a GPU for better GPU utility.

Selecting Data Sources

To easily bring your code and datasets into your workspace, you can select one or more data sources. This allows you to access your code from a Git repository and your dataset from an NFS (Network File System).

Creating the Workspace

After selecting the necessary tools, versions, compute resources, and data sources, click on "Create Workspace" to create your new workspace. With Run AI Workspaces, you no longer need to worry about building container images or configuring ingress settings. You can now focus on your research and easily turn your environments on and off when needed.

Connecting to a Workspace

Once you have created a workspace, you can connect to it and start working on your projects.

Accessing Workspace Tools

To access the tools assigned to your workspace, click on the "Connect" button. This will Show you the available tools, such as Jupiter Notebook or MLflow, depending on your workspace configuration.

Working with Jupiter Notebook

If you choose Jupiter Notebook as your tool, clicking on "Jupiter" will automatically launch Jupiter Lab Notebook for your workspace. You can then start working on your models, experimenting, and analyzing data within the Jupiter environment.

Using MLflow

Similarly, if you have assigned MLflow to your workspace, clicking on "MLflow" will give you access to the MLflow UI. This allows you to track experiments, Package and deploy models, and collaborate with other team members.

Managing Workspaces

Run AI Workspaces provides convenient options for managing your workspaces.

Starting and Stopping Workspaces

You have the flexibility to start and stop your workspaces as needed. When you are finished working in your environment, simply click on "Stop" to stop your workspace. This frees up the compute resources assigned to your workspace, allowing other researchers to utilize them.

Sharing Compute Resources

Run AI Workspaces also allows sharing compute resources. This means that if you have unused GPU resources, other researchers can utilize them to enhance GPU utility station, maximizing resource utilization within the organization.

Customizing Workspaces

With Run AI Workspaces, you have the ability to customize and tailor your workspaces according to your requirements.

Creating Templates

You can create templates to define the environment, data source, and compute resources. These templates can be reused for multiple workspaces, saving time and effort in setting up each workspace individually.

Editing Environment Settings

To modify the tools, versions, or frameworks in your workspace, you can easily edit the environment settings. This enables you to stay up-to-date with the latest tools and libraries without the need to start from scratch.

Modifying Compute Resources

If your resource requirements change, you can modify the compute resources assigned to your workspace. Whether it's scaling up for more intensive workloads or scaling down to conserve resources, you have the flexibility to adapt your workspace to your needs.

Conclusion

Run AI Workspaces simplifies the process of creating and managing work environments for data scientists. By providing a user-friendly interface and seamless integration with tools and data sources, it allows researchers to focus on their research, increase productivity, and optimize resource utilization. With the flexibility to start, stop, customize, and share workspaces, Run AI Workspaces offers an efficient solution for data scientists to accelerate their work.

Highlights

• Run AI Workspaces provides a simplified way for data scientists to create and manage their work environments.
• Users can easily choose their preferred tools, versions, and compute resources without the hassle of building container images or configuring networking and storage.
• Workspaces can be started and stopped as needed, allowing for efficient resource utilization.
• Sharing compute resources enhances GPU utility station and maximizes resource utilization within the organization.
• Customizable templates and environment settings enable users to tailor their workspaces according to their requirements.
• Run AI Workspaces helps data scientists focus on their research, increase productivity, and optimize resource management.

FAQ

Q: How does Run AI Workspaces simplify the process of creating work environments? A: Run AI Workspaces eliminates the need to build container images or configure networking and storage, allowing data scientists to focus on their research.

Q: Can I customize my workspace according to my requirements? A: Yes, Run AI Workspaces provides options to customize workspaces by selecting different tools, versions, compute resources, and data sources.

Q: Can I share compute resources with other researchers? A: Yes, Run AI Workspaces allows sharing compute resources, maximizing resource utilization within your organization.

Q: Can I start and stop my workspaces as needed? A: Yes, you have the flexibility to start and stop your workspaces, conserving resources when not in use.

Q: Can I modify my workspace's compute resources? A: Yes, you can easily modify the compute resources assigned to your workspace based on your changing requirements.