Merge pull request #69 from deep-diver/firebase-proposal

Project proposal for Firebase ML Publisher custom component

proposals/20210817-firebase_ml_publisher_component.md

@@ -0,0 +1,46 @@
+#### SIG TFX-Addons
+# Project Proposal
+
+**Your name:** Chansung Park
+
+**Your email:** [email protected]
+
+**Your company/organization:** Individual(ML GDE)
+
+**Project name:** [Firebase ML Publisher](https://github.com/tensorflow/tfx-addons/issues/59)
+
+## Project Description
+This project defines a custom TFX component to publish/update ML models to [Firebase ML](https://firebase.google.com/products/ml). This is another type of pusher component, and the input model is assumed to be a TFLite format.
+
+## Project Category
+Component
+
+## Project Use-Case(s)
+This project helps users to publish trained models directly to Firebase ML.
+
+With Firebase ML, we can guarantee that mobile devices can be equipped with the latest ML model without explicitly embedding binary in the project compiling stage. We can even A/B test different versions of a model with Google Analytics when the model is published on Firebase ML.
+
+## Project Implementation
+Firebase ML Publisher component will be implemented as Python function-based component. You can find the [actual source code](https://github.com/sayakpaul/Dual-Deployments-on-Vertex-AI/blob/main/custom_components/firebase_publisher.py) in my personal project. Please note this is a personal implementation, and it will be enhanced as a official TFX Addon component.
+
+The implementation details
+- Define a custom Python function-based TFX component. It takes the following parameters from a previous component.
+  - It should follow the standard Pusher's interface since this is another custom pusher.
+  - Additionally, it takes meta information to manage published model for Firebase ML such as `display name` and `tags`.
+- Download saved TFLite model file by referencing the output from a previous component
+  - Firebase SDK doesn't allow to publish models from GCS directly.
+- Initialize Firebase Admin with the credential and Firebase temporary-use GCS bucket.
+  - Firebase credentials can be setup via [Workload Identity](https://cloud.google.com/kubernetes-engine/docs/how-to/workload-identity) for GKE or [Mounting Secret API in TFX runner](https://github.com/tensorflow/tfx/blob/d989bbd7fc366c73ad833428ce6b5cf57a587432/tfx/orchestration/kubeflow/kubeflow_dag_runner.py#L78).
+- Search if any models with the same `display name` has already been published.
+  - if yes, update the existing Firebase ML mode, then publish it
+  - if no, create a new Firebase ML model, then publish it
+- Return `tfx.dsl.components.OutputDict` to indicate if the job went successful, and if the job was about creating a new Firebase ML model or updating the exisitng Firebase ML model.
+
+## Project Dependencies
+The implementation will use the following libraries.
+- [Firebase Admin Python SDK](https://github.com/firebase/firebase-admin-python) >= 5.0.2
+- [Python Client for Google Cloud Storage](https://github.com/googleapis/python-storage) >= 1.42.0
+
+## Project Team
+**Project Leader** : Chansung Park, deep-diver, [email protected]
+1. Sayak Paul, sayakpaul, [email protected]