Name of the project: ForestFlow
Requested maturity level: Incubation
Description
ForestFlow is a scalable policy-based cloud-native machine learning model server. ForestFlow strives to strike a balance between the flexibility it offers data scientists and the adoption of standards while reducing friction between Data Science, Engineering and Operations teams.
ForestFlow is policy-based because we believe automation for Machine Learning/Deep Learning operations is critical to scaling human resources. ForestFlow lends itself well to workflows based on automatic retraining, version control, A/B testing, Canary Model deployments, Shadow testing, automatic time or performance-based model deprecation and time or performance-based model routing in real-time.
Our aim with ForestFlow is to provide data scientists a simple means to deploy models to a production system with minimal friction accelerating the development to production value proposition.
To achieve these goals, ForestFlow looks to address the proliferation of model serving formats and standards for inference API specifications by adopting, what we believe, are currently, or are becoming widely adopted open source frameworks, formats, and API specifications. We do this in a pluggable format such that we can continue to evolve ForestFlow as the industry and space matures and we see a need for additional support.
Alignment with LF AI’s mission:
ForestFlow provides the community with an easy way to deploy ML models to production and realize business value on an open-source platform that can scale as user’s projects and requirements scale.
Possible integrations with existing LF AI projects:
https://github.com/onnx/onnx - ONNX is an obvious choice to integrate with ForestFlow as a flavor of models that ForestFlow supports deployments for.
License: Apache License 2.0 https://github.com/dreamworksanimation/ForestFlow/blob/master/LICENSE
Source control: https://github.com/dreamworksanimation/ForestFlow
Collaboration tools: None currently in use. Need mailing list and slack channel for users, developers and TSC.
External dependencies & licenses: Please refer to: https://github.com/dreamworksanimation/ForestFlow/blob/master/NOTICE.md
Initial committers:
-
Ahmad Alkilani ahmad.alkilani@dreamworks.com, since inception
Project role definitions: None yet, except DreamWorks has a Contribution guide and CLA for corporations and individuals. I imagine this would be supplanted by LFAI if project moves over. See: https://github.com/dreamworksanimation/ForestFlow/blob/master/docs/contributing.md
Total # of contributors & affiliations: 3 total contributors all affiliated with DreamWorks Animation.
Release methodology:
Not fully fledged yet however a release is performed when committers pass the vote to do so. Preferred release approach is via a docker image. Please see requirements for dockerhub. CI also needs to be setup in addition to test-case coverage which is minimal at the moment. An updated RELEASES.md will be provided upon project approval and governance.
Code of conduct: Nothing exists today outside of CLAs, see here: https://github.com/dreamworksanimation/ForestFlow/blob/master/docs/contributing.md Creation of CODE_OF_CONDUCT.md file that points to LF CoC will be created upon project approval and transition of Github.
Infrastructure requests:
We’ll need some sort of CI in addition to a dockerhub account to provide official release images
Project Website: Domain is not registered. Relying on Github pages generated from docs at the moment: https://dreamworksanimation.github.io/ForestFlow/
We will request design resources to design a proper website.
Project governance: None.
Social media accounts: None
Existing sponsorship: DreamWorks Animation has previously provided developer resources to create and improve upon ForestFlow.