Demonstration of a project structure for common machine learning workflows. Uses Tensorflow and Scikit with a monorepo structure.
An example of executing via REST (FastAPI) is also included.
$ pyenv virtualenv 3.11.9 tf-sk-workflows$ pyenv local tf-sk-workflowspoetry installcore
Core classes and abstractions related to all models, representations of model performance and comparison utils.
sklearnworkflows
Machine learning pipelines implemented using Sklearn.
tensorflowworkflows
Machine learning pipelines implemented using Tensorflow.
api A simple FastAPI interface that exposes some API endpoints to demonstrate executing the workflows via REST.
poetry run python apiAvailable at:
SwaggerUI: http://localhost:3001/docs
Redoc: http://localhost:3001/redoc
JSON Spec: http://localhost:3001/openapi.json