Skip to main content

mlflow: An experiment framework

Project description


MLflow requires conda to be on the PATH for the projects feature.

We recommend installing MLflow in its own virtualenv for development, as follows:

virtualenv env
source env/bin/activate
pip install -r dev-requirements.txt
pip install -r tox-requirements.txt
pip install -e .

Running a Sample App With the Tracking API

The programs in example use the MLflow Tracking API. For instance, run:

python example/

This program will use MLflow log API, which stores tracking data in ./mlruns, which can then be viewed with the Tracking UI.

Launching the Tracking UI

Development Mode

If you installed MLflow for development (e.g. via pip install -e .), npm is required to build and run the UI. You can verify that npm is on the PATH by running npm -v, and install npm if needed.

We recommend Running the Javascript Dev Server - otherwise, the tracking frontend will request files in the mlflow/server/js/build directory, which is not checked into Git. Alternatively, you can generate the necessary files in mlflow/server/js/build as described in Building a Distributable Artifact.

Starting the UI

The MLflow Tracking UI will show runs logged in ./mlruns at http://localhost:5000. Start it with:

mlflow ui

Running a Project from a URI

The mlflow run command lets you run a project packaged with a MLproject file from a local path or a Git URI:

mlflow run example/project -P num_dimensions=5

mlflow run -P num_dimensions=5

See example/project for a sample project with an MLproject file.

Saving and Serving Models

To illustrate managing models, the mlflow.sklearn package can log Scikit-learn models as MLflow artifacts and then load them again for serving. There is an example training application in example/ that you can run as follows:

$ python example/
Score: 0.666
Model saved in run RUN_ID

$ mlflow sklearn serve -r RUN_ID model

$ curl -d '[{"x": 1}, {"x": -1}]' -H 'Content-Type: application/json' -X POST localhost:5000/invocations

Tests and Lint


Install Node Modules

Before running the Javascript dev server or building a distributable wheel, install Javascript dependencies via:

cd mlflow/server/js
npm install
cd - # go back to root repository directory

If modifying dependencies in mlflow/server/js/package.json, run npm update within mlflow/server/js to install the updated dependencies.

Running the Javascript Dev Server

Install Node Modules, then run the following:

In one shell:

mlflow ui

In another shell:

cd mlflow/server/js
npm start

The MLflow Tracking UI will show runs logged in ./mlruns at http://localhost:3000.

Building a Distributable Artifact

Install Node Modules, then run the following:

Generate JS files in mlflow/server/js/build:

cd mlflow/server/js
npm run build

Build a pip-installable wheel in dist/:

cd -
python bdist_wheel

Writing Docs

Make sure you have the requirements from dev-requirements.txt installed. Then run

cd docs
make livehtml

Now Go With the Flow!

Build Status

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for mlflow, version 0.0.1
Filename, size File type Python version Upload date Hashes
Filename, size mlflow-0.0.1-py2-none-any.whl (4.3 MB) File type Wheel Python version 2.7 Upload date Hashes View hashes
Filename, size mlflow-0.0.1.tar.gz (4.3 MB) File type Source Python version None Upload date Hashes View hashes

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN SignalFx SignalFx Supporter DigiCert DigiCert EV certificate StatusPage StatusPage Status page