Ray

Ray is a framework for scaling apps based on AI and Python.

Ray includes a distributed runtime environment and AI libraries that simplify ML computations. This framework is capable of processing large amount of data with high CPU and GPU utilization. You can run Ray on a single VM and then scale it up to a cluster, thus abstracting distributed computations.

1. Get an SSH key pair to connect to a virtual machine.

2. Create a VM from a public image. Under Image/boot disk selection, go to the Cloud Marketplace tab and select Ray. Under Access:

   • Enter the username in the Login field.
   • Paste the contents of the public SSH key file in the SSH key field.

   Save the VM public IP address.

3. Connect to the VM over SSH. Use the username you set when creating the VM and the private SSH key you created before.

4. To create SSH tunnels and connect to client-server, Global Control Store (GCS), or Ray control panel, use the following commands:

   # Client-server
   ssh -L 10001:localhost:10001 <username>@<VM_public_IP_address>
   

   # GCS
   ssh -L 6379:localhost:6379 <username>@<VM_public_IP_address>
   

   # Dashboard
   ssh -L 8265:localhost:8265 <username>@<VM_public_IP_address>
   

5. To open Ray control panel in your browser, go to http://localhost:8265.

• Scalable ML data processing.
• Using ML models for batch inference.
• Collecting data for ML model training.
• Pre-processing data for ML model training.
• ML model distributed training.
• Fine-tuning ML models based on new data.
• Configuring ML model hyperparameters.
• Reinforcement learning.
• Developing APIs to work with trained ML models.

Yandex Cloud technical support is available 24/7. The types of requests you can submit and the relevant response times depend on your pricing plan. You can switch to the paid support plan in the management console. You can learn more about the technical support terms and conditions here.