Welcome to FloydHub! Here you'll find comprehensive information for training and deploying your deep learning and AI applications with our platform. We do our best to make this documentation clear and user friendly, but if you have unanswered questions, please visit the community forum or contact us.

The fastest way to get up and running is to use our quickstart guide, which walks through an entire FloydHub training job step-by-step. You'll create a new Project on the FloydHub web dashboard, connect it to a local directory on your computer, and then kick-off a job using the FloydHub CLI to train your deep learning model on FloydHub's GPU servers.

Deep learning without the DevOps:

Frictionless data science

Why worry about provisioning GPUs, installing drivers, or managing software dependency hell? With FloydHub, we take care of your entire deep learning DevOps workflow - so you can focus on the science.

Training a TensorFlow model using GPUs on the cloud is as simple as executing this command on your terminal: floyd run --gpu --env tensorflow "python train.py". Try it now with our quickstart guide.

Powerful workflow tools

Whether you're using our web dashboard or our command line interface, our tools make your work easier and your team more productive:

Productivity for teams

FloydHub's Team Plan is a collaborative tool for data science teams of all sizes:

  • Centralized, secure hub for all your team's machine learning experiments
  • Team management with roles-based permissioning
  • Consolidated billing and usage tracking
  • High priority customer support

Deep learning community

FloydHub hosts open source Projects and Datasets that you can discover, clone, and reproduce (or reconfigure with your own Dataset).

Try it now with the Neural Style Transfer.

We're here to help!

We're always happy to help with any questions you might have! Search our documentation or check answers to frequently asked questions. The FloydHub community forum is another place to ask questions, request features, or share cool Projects. For more help, send us an email.