mlflow: [BUG] ServiceWorker tries to precache assets from /static-files/static-files instead of /static-files

System information

• Have I written custom code (as opposed to using a stock example script provided in MLflow): no
• OS Platform and Distribution (e.g., Linux Ubuntu 16.04): Linux
• MLflow installed from (source or binary): using pip install mlflow
• MLflow version (run mlflow --version): 1.2.0
• Python version: 3.6.6
• npm version, if running the dev UI: N/A
• Exact command to reproduce: mlflow server --backend-store-uri sqlite:////tmp/mlflow.db --default-artifact-root file:////tmp/ --host 0.0.0.0

Describe the problem

The serviceworker is issuing request to /static-files/static-files/somefile.css instead of /static-files/somefile.css. This cause the initial load to be very slow especially with the default 4 gunicorn workers.

For instance, requests to http://localhost:5000/asdf/static-files/static-files/static/css/main.708e4447.css fails while http://localhost:5000/asdf/static-files/static/css/main.708e4447.css` would work.