Start Here
Watch the brief video and do the hands-on tutorial below to quickly onboard and become productive with your ML research on the Nightingale platform.
Getting Started
- Register for the Nightingale OS Platform.
- Wait to get admitted; we’ll onboard users as quickly as possible.
- Log on to the platform, create a new project
- Create a new instance within the project to explore the data, and to train and test your models.
- Use cpu.xsmall instances to explore the data and train/test your models for free, or
- Fund a project using a credit card to use our full range of instance sizes, including GPUs.
- Do the hands-on tutorial below to learn to access the data, train and test ML models quickly on our platform.
- Invite collaborators to join your Nightingale project.
For more details about getting started, refer to our platform docs.
Setting up a deep learning model
In this video, William demonstrates how to access our platform and start an instance. He then runs a notebook that trains a model to predict breast cancer stage from the downsampled version of the brca-psj-path breast cancer biopsy images dataset using PyTorch. Finally, he tests the accuracy of the deep neural net (resnet) model.
Tutorial Notebook
An executable version of this notebook can be found on our platform at ~/datasets/_start_here/. Please open this within your instance using JupyterLab to do the hands-on tutorial.