New Component: trubrics, collect human feedback on your ML models in Streamlit

[jeffkayne]

Hey all,

At Trubrics we are building tools to help you enhance your ML workflows with human feedback.

We’ve released a new component with Streamlit that allows you to collect qualitative and quantitative feedback from your users and save it along with all your model and dataset metadata:

Install it with:

pip install "trubrics[streamlit]"

and start collecting feedback in your app with:

from trubrics.integrations.streamlit import FeedbackCollector

collector = FeedbackCollector()
collector.st_feedback(feedback_type="issue")

Demo: https://trubrics-titanic-example.streamlit.app/

Source code: GitHub - trubrics/trubrics-sdk: Enhance your ML workflows with human feedback

Thanks & please let us know what you think !

[Caroline]

Awesome!

[Hemanth_Sai]

The component is so good. But this works on an older version of Streamlit.

[jeffkayne]

Hello @Hemanth_Sai

Thanks for pointing this out, we’ve fixed it in the new release! Have you given it a try? Please don’t hesitate if you have any feedback !

Best
Jeff

[Hemanth_Sai]

Sorry if I’m pointing out too many version compatibilities😅 but the component installs scikit-learn 1.0.2 but the current scikit version is 1.2.2.

[jeffkayne]

True, I’ll sort this out in the next release.

For the moment don’t worry about the scikit-learn version required, as it only concerns another part of the package. Just run a pip install --upgrade scikit-learn and all should work just fine

[Hemanth_Sai]

Used trubrics feedback component. I need to connect it to the database to store the feedback (cosmosDB) and need to do some UI changes. Here’s the Link: Termsbrief

[jeffkayne]

Awesome!

Checkout out st.form() to wrap your custom feedback in a form