[FR] "Freeze" experiments in ui to static html site

Thank you for submitting a feature request. Before proceeding, please review MLflow’s Issue Policy for feature requests and the MLflow Contributing Guide.

Please fill in this feature request template to ensure a timely and thorough response.

Willingness to contribute

The MLflow Community encourages new feature contributions. Would you or another member of your organization be willing to contribute an implementation of this feature (either as an MLflow Plugin or an enhancement to the MLflow code base)?

• Yes. I can contribute this feature independently.
• Yes. I would be willing to contribute this feature with guidance from the MLflow community.
• No. I cannot contribute this feature at this time.

Proposal Summary

I’d like to dump mlflow’s ui into a static site with a few experiments, to be hosted at gh-pages or a similar service. These experiments are the accompanying material to a research paper. No new experiments will be added, and no changes will be done; so doing one “dump” of the ui interface into plain HTML/CSS/JS would solve this.

I couldn’t find any previous posts about this.

Motivation

• What is the use case for this feature?

I have a set of experiments accompanying an upcoming paper. There will be no more runs for this project, and I want to share all the results of the experiments I did. I find the mlflow ui very easy to use and, ideally, I’d just like to deploy the relevant experiments in a gh-pages or similar site.

• Why is this use case valuable to support for MLflow users in general?

It might be relevant for people who want to share the interactive visualization of their experiments without too much hazzle or reinventing a website from scratch.

• Why is it currently difficult to achieve this use case? (please be as specific as possible about why related MLflow features and components are insufficient)

It is my impression that it shouldn’t be that difficult to do, assuming I already have everything in my mlruns folder and that there will be no further experiments for this project. I essentially want to “freeze” these visualizations permanently as a supplementary material of my research paper.

Maybe something like Frozen-Flask will do? But I’d be very grateful if I can get some help on this.

What component(s), interfaces, languages, and integrations does this feature affect?

Components

• area/artifacts: Artifact stores and artifact logging
• area/build: Build and test infrastructure for MLflow
• area/docs: MLflow documentation pages
• area/examples: Example code
• area/model-registry: Model Registry service, APIs, and the fluent client calls for Model Registry
• area/models: MLmodel format, model serialization/deserialization, flavors
• area/projects: MLproject format, project running backends
• area/scoring: MLflow Model server, model deployment tools, Spark UDFs
• area/server-infra: MLflow Tracking server backend
• area/tracking: Tracking Service, tracking client APIs, autologging

Details

Any guidance on how to achieve this would be greatly appreciated.

I already tried scraping my localhost:5000 but I keep getting error messages. Maybe ajax operations going on? I understand if it’s unfeasible to do this, but I thought that what I want to store is really those HTML files with concluded experiments. Maybe doable?

Thanks.

(Use this section to include any additional information about the feature. If you have a proposal for how to implement this feature, please include it here. For implementation guidelines, please refer to the Contributing Guide.)

[matiasdahl]

This may be relevant

• https://pynb-dag-runner.github.io/mnist-digits-demo-pipeline (demo site takes a while to load)

This is built using a (modified version of) MLFlow that outputs a read only static site that can be deployed to Github Pages. Ie., the above site does not depend on a backend or API.

This uses a pynb-dag-runner library to run a demo pipelines on Github Action compute resources with reporting as a Github page. This is WIP.

[napulen]

@matiasdahl That’s awesome, virtually the same thing I’ve tried to do. I’ll take a look into the setup of this gh-page demo.