Open Source
Floyd is an open source project. We'd love for you to get involved
Beginner Friendly
Easy to use. Plenty of plug and play modules. Get up and running in no time
Tools for Experts
Floyd automates the common tasks so you can focus on deep learning
Interactive, Visual Workspace
Drag and drop datasets and modules to build your pipeline visually
- Easy to build and monitor complex pipelines
- Create your own modules, or choose from popular community modules
- Floyd manages environments and dependencies using Docker, so you don't have to
- Intelligent caching to avoid rerunning expensive modules
Iterate on you Ideas, hassle free
Floyd keeps your code, data, parameters, results and environments in one place
- Run multiple experiments in parallel and keep everything tracked
- Make all your experiments fully reproducible and shareable
- Floyd automated scheduling, notifications & resource monitoring
- Run in the cloud or on-prem and integrate with your infrastructure
Features
-
CLI and Python SDK
Not a fan of GUI? Use Floyd via CLI, REST API or Python SDK to integrate into your workflow
-
Registry of plug-and-play Modules
Use popular deep learning algorithms from the registy in your pipeline. Batteries included
-
Plays well with Others
Floyd integrates seamlessly with Docker, Github and AWS. More coming soon!
Floyd Cloud
(Coming soon) A central hub to run your deep learning pipelines on scalable GPU hardware, store your data and collaborate in realtime
- Provision GPUs with zero-setups
- Ship models to production in one click
- Real time collaboration