Steps

Make front end changes

1. Adjust App.css and App.js

Build pip wheel

1. Fulfill reqs as described here https://github.com/mlflow/mlflow/blob/master/CONTRIBUTING.rst#common-prerequisites-and-dependencies conda create --name mlflow-dev-env python=3.6 pip install -r dev-requirements.txt pip install -r test-requirements.txt pip install -e .[extras] # installs mlflow from current checkout pip install pytest
2. Also Javascript dependencies cd mlflow/server/js npm install cd - # return to root repository directory
3. Build a dstributable artifcat https://github.com/mlflow/mlflow/blob/master/CONTRIBUTING.rst#building-a-distributable-artifact
   1. Generate JS files in mlflow/server/js/build cd mlflow/server/js npm run build
   2. Build a pip-installable wheel in dist/: cd - python setup.py bdist_wheel

Install pip wheel (it is in /dist)

pip install file.whl

Run mlflow as usual

mlflow server

Troube shooting

1. Update pandas pip install pandas -U
2. Cannot find module 'caniuse-lite/dist/unpacker/agents' when running create react app npm i caniuse-lite

Inside aws mlflow cloud formation

1. Adjust Dockerfile to wget and install custom wheel
2. Rest as normal

regular aws-mlflow: https://github.com/aws-samples/amazon-sagemaker-mlflow-fargate

project based mlflow example with wheel: https://github.com/aiaudit-org/mlflow-pm

itu aws-mlflow: https://github.com/aiaudit-org/amazon-sagemaker-mlflow-fargate

itu aws-mlflow w/ project based mflow wheel: https://github.com/aiaudit-org/mlflow-aws-pm