MLSync

MLSync is a productivity tool that syncs your ML data with productivity tools you love.

Installation

pip install mlsync

Example: MLFlow -> Notion

Sync your machine learning experiments to Notion in a few simple steps!

Configuration Setup

Let us first setup the run environment.

  1. To begin, checkout this repository: git clone https://github.com/paletteml/mlsync.git
  2. Change to the mlsync/examples/mlflow-notion directory: cd mlsync/examples/mlflow-notion/
  3. Rename the .env.example file in your path: mv .env.example .env. This file is intended to store your personal API keys.

Note that the directory contains YAML files for configurations (config.yaml) and report formatting (format.yaml). We will leave the configurations as is for now.

ML Training Setup

Now let us setup our ML Training environment. For this example, we will rely on the MLFlow framework and Pytorch as our ML framework. Since MLFlow supports all major ML frameworks, this example can be easily adapted to other frameworks.

  1. If not already installed, install PyTorch based on the guide here. (Only needed for the provided example).
  2. Install mlflow package using pip install mlflow. More about installation here.
  3. Run example training using python mlflow_pytorch.py --run-name <Run 1>. This will create a new MLFlow run.
  4. Launch MLFlow UI using mlflow ui &. Copy the mlflow uri (seen in the command line as [INFO] Listening at: <URL>). Let it run in the background.
  5. Update the uri field in the configuration file in your folder (config.yaml) under mlflow with the just copied mlflow uri.

Notion Setup

Let us now link Notion to MLSync. This is required only for the first time you run MLSync.

  1. Create a new integration to Notion.
    1. Visit notion.so/my-integrations
    2. Click the + New Integration button
    3. Let us name it as MLSync.
    4. Ensure Read, Update and Insert Content Capabilities are selected.
    5. Copy your “Internal Integration Token” from your Notion integration page into the .env file in your path.
      • NOTION_TOKEN=secret_0000000000000000000000000000000000000000000
  2. Create a new page in Notion. This will serve as the root page for your MLFlow runs.
    1. Click Share button on the top right corner of the page.
    2. Click Invite button and then choose MLSync integration.
    3. Back in the Share dialog, click the Copy link button.
    4. Paste the URL to the page_id field in the configuration file (config.yaml) under notion.

Syncing

You are now all set! Now let us sync your MLFlow runs to Notion.

mlsync --config config.yaml --report_format format.yaml

That’s it! You can now view your MLFlow runs in Notion. As long as mlsync is running, all your future experiments and runs should appear in selected Notion page.

Advanced

  1. You can override the Notion page id, token, and other configurations by either modifying the config.yaml file or by passing the arguments to the mlsync command. Run mlsync --help to see the available arguments.
  2. Custom Report Formats: mlsync allows you to customize the report much further. You can customize the report by adding your own format.yaml file. Read documentation here to learn more.
  3. Custom Refresh Rates: You can control the refresh rate of the report by setting the refresh_rate field in the configuration file.
  4. Restarting mlsync: You can restart mlsync any time without losing earlier runs.

Enjoy! If you have any further questions, please contact us.

Roadmap

We want to support different training enviroments and different productivty tools.

  1. Productivity Tools
    1. Notion: Supported
    2. Trello: Planned
    3. Confluence: In progress
    4. Jira: Planned
  2. Monitoring Frameworks
    1. MLFlow: Supported
    2. TensorBoard: Planned
    3. ClearML: Planned

Do you have other tools/frameworks you would like to see supported? Let us know!

Contributing

We welcome contributions from the community. Please feel free to open an issue or pull request. Or, if you are interested in working closely with us, please contact us directly. We will be happy to talk with you!

GitHub

View Github