Data Science Virtual Machine

Data Science Virtual Machine (DSVM) is an Ubuntu Server-based cloud server optimized for data science and machine learning projects.

The VM image includes the most popular ML tools like PyTorch, TensorFlow, Jupyter Notebook and lets you work on your data science project out of the box.

Use Cases

Data Science Virtual Machine can be used for:

• application development:
  • chat bots;
  • recommendation services;
  • object recognition (in images or videos);
  • speech recognition and synthesis;
  • forecasting services;
• model training;
• data experiments.

Hardware Configuration

Minimal configuration is 2 GB RAM and 25 GB disk space.

For faster model training we recommend using the GPU Line configurations that include up to 4 GPUs.

Pre-installed Tools

• Python 3.6 with the following packages:
  • PIP
  • PyTorch
  • TensorFlow
  • JupyterLab
  • Keras
  • Jupyter Notebook
  • scikit-learn
  • Numpy
  • Scipy
  • Pandas
  • NLTK
  • OpenCV
  • Catboost
  • XGBoost
  • LightGBM

Creating New Servers

Follow these steps to create a new instance:

1. Log in to the Control panel.
2. Go to the Cloud Platform section.
3. Click Create server.
4. Click Choose another source in the Source section.
5. Choose Ubuntu 18.04 LTS Machine Learning 64-bit under the Default images tab.
6. Click Select.
7. Configure the virtual machine according to the instruction.
8. Click Create to complete the setup.

JupyterLab Quickstart

Type http://[cloud-server-ip-address] in your browser to launch JupyterLab. The default password is: ikieg2wahmohtahF.

Billing and Payment

Data Science Virtual Machine runs on Cloud Platform, so the funds are withdrawn using the pay-as-you-go model according to the Cloud Platform pricing. Payment for used resources is performed hourly. Funds are withdrawn for the entire existence time of the Virtual Machine, regardless of whether it is turned on or off.