MLSync is a productivity tool that syncs your ML data with productivity tools you love.
pip install mlsync
Example: MLFlow -> Notion
Sync your machine learning experiments to Notion in a few simple steps!
Let us first setup the run environment.
- To begin, checkout this repository:
git clone https://github.com/paletteml/mlsync.git
- Change to the
- Rename the
.env.examplefile 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.
- If not already installed, install PyTorch based on the guide here. (Only needed for the provided example).
pip install mlflow. More about installation here.
- Run example training using
python mlflow_pytorch.py --run-name <Run 1>. This will create a new MLFlow run.
- 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.
- Update the
urifield in the configuration file in your folder (
mlflowwith the just copied mlflow uri.
Let us now link Notion to MLSync. This is required only for the first time you run MLSync.
- Create a new integration to Notion.
- Visit notion.so/my-integrations
- Click the
+ New Integrationbutton
- Let us name it as
InsertContent Capabilities are selected.
- Copy your “Internal Integration Token” from your Notion integration page into the
.envfile in your path.
- Create a new page in Notion. This will serve as the root page for your MLFlow runs.
- Click Share button on the top right corner of the page.
- Click Invite button and then choose
- Back in the
Sharedialog, click the
- Paste the URL to the
page_idfield in the configuration file (
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.
- You can override the Notion page id, token, and other configurations by either modifying the
config.yamlfile or by passing the arguments to the
mlsync --helpto see the available arguments.
- Custom Report Formats:
mlsyncallows you to customize the report much further. You can customize the report by adding your own
format.yamlfile. Read documentation here to learn more.
- Custom Refresh Rates: You can control the refresh rate of the report by setting the
refresh_ratefield in the configuration file.
- Restarting mlsync: You can restart mlsync any time without losing earlier runs.
Enjoy! If you have any further questions, please contact us.
We want to support different training enviroments and different productivty tools.
- Productivity Tools
- Monitoring Frameworks
Do you have other tools/frameworks you would like to see supported? Let us know!
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!